feat: RESP3 notifications support & Hitless notifications handling [CAE-1088] & [CAE-1072] (#3418)

- Adds support for handling push notifications with RESP3. 
- Using this support adds handlers for hitless upgrades.

---------

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Hristo Temelski <hristo.temelski@redis.com>

.gitignore

@@ -9,3 +9,6 @@ coverage.txt
 **/coverage.txt
 .vscode
 tmp/*
+
+# Hitless upgrade documentation (temporary)
+hitless/docs/

adapters.go

@@ -0,0 +1,111 @@
+package redis
+
+import (
+	"context"
+	"errors"
+	"net"
+	"time"
+
+	"github.com/redis/go-redis/v9/internal/interfaces"
+	"github.com/redis/go-redis/v9/push"
+)
+
+// ErrInvalidCommand is returned when an invalid command is passed to ExecuteCommand.
+var ErrInvalidCommand = errors.New("invalid command type")
+
+// ErrInvalidPool is returned when the pool type is not supported.
+var ErrInvalidPool = errors.New("invalid pool type")
+
+// newClientAdapter creates a new client adapter for regular Redis clients.
+func newClientAdapter(client *baseClient) interfaces.ClientInterface {
+	return &clientAdapter{client: client}
+}
+
+// clientAdapter adapts a Redis client to implement interfaces.ClientInterface.
+type clientAdapter struct {
+	client *baseClient
+}
+
+// GetOptions returns the client options.
+func (ca *clientAdapter) GetOptions() interfaces.OptionsInterface {
+	return &optionsAdapter{options: ca.client.opt}
+}
+
+// GetPushProcessor returns the client's push notification processor.
+func (ca *clientAdapter) GetPushProcessor() interfaces.NotificationProcessor {
+	return &pushProcessorAdapter{processor: ca.client.pushProcessor}
+}
+
+// optionsAdapter adapts Redis options to implement interfaces.OptionsInterface.
+type optionsAdapter struct {
+	options *Options
+}
+
+// GetReadTimeout returns the read timeout.
+func (oa *optionsAdapter) GetReadTimeout() time.Duration {
+	return oa.options.ReadTimeout
+}
+
+// GetWriteTimeout returns the write timeout.
+func (oa *optionsAdapter) GetWriteTimeout() time.Duration {
+	return oa.options.WriteTimeout
+}
+
+// GetNetwork returns the network type.
+func (oa *optionsAdapter) GetNetwork() string {
+	return oa.options.Network
+}
+
+// GetAddr returns the connection address.
+func (oa *optionsAdapter) GetAddr() string {
+	return oa.options.Addr
+}
+
+// IsTLSEnabled returns true if TLS is enabled.
+func (oa *optionsAdapter) IsTLSEnabled() bool {
+	return oa.options.TLSConfig != nil
+}
+
+// GetProtocol returns the protocol version.
+func (oa *optionsAdapter) GetProtocol() int {
+	return oa.options.Protocol
+}
+
+// GetPoolSize returns the connection pool size.
+func (oa *optionsAdapter) GetPoolSize() int {
+	return oa.options.PoolSize
+}
+
+// NewDialer returns a new dialer function for the connection.
+func (oa *optionsAdapter) NewDialer() func(context.Context) (net.Conn, error) {
+	baseDialer := oa.options.NewDialer()
+	return func(ctx context.Context) (net.Conn, error) {
+		// Extract network and address from the options
+		network := oa.options.Network
+		addr := oa.options.Addr
+		return baseDialer(ctx, network, addr)
+	}
+}
+
+// pushProcessorAdapter adapts a push.NotificationProcessor to implement interfaces.NotificationProcessor.
+type pushProcessorAdapter struct {
+	processor push.NotificationProcessor
+}
+
+// RegisterHandler registers a handler for a specific push notification name.
+func (ppa *pushProcessorAdapter) RegisterHandler(pushNotificationName string, handler interface{}, protected bool) error {
+	if pushHandler, ok := handler.(push.NotificationHandler); ok {
+		return ppa.processor.RegisterHandler(pushNotificationName, pushHandler, protected)
+	}
+	return errors.New("handler must implement push.NotificationHandler")
+}
+
+// UnregisterHandler removes a handler for a specific push notification name.
+func (ppa *pushProcessorAdapter) UnregisterHandler(pushNotificationName string) error {
+	return ppa.processor.UnregisterHandler(pushNotificationName)
+}
+
+// GetHandler returns the handler for a specific push notification name.
+func (ppa *pushProcessorAdapter) GetHandler(pushNotificationName string) interface{} {
+	return ppa.processor.GetHandler(pushNotificationName)
+}

async_handoff_integration_test.go

@@ -0,0 +1,353 @@
+package redis
+
+import (
+	"context"
+	"net"
+	"sync"
+	"testing"
+	"time"
+
+	"github.com/redis/go-redis/v9/hitless"
+	"github.com/redis/go-redis/v9/internal/pool"
+	"github.com/redis/go-redis/v9/logging"
+)
+
+// mockNetConn implements net.Conn for testing
+type mockNetConn struct {
+	addr string
+}
+
+func (m *mockNetConn) Read(b []byte) (n int, err error)   { return 0, nil }
+func (m *mockNetConn) Write(b []byte) (n int, err error)  { return len(b), nil }
+func (m *mockNetConn) Close() error                       { return nil }
+func (m *mockNetConn) LocalAddr() net.Addr                { return &mockAddr{m.addr} }
+func (m *mockNetConn) RemoteAddr() net.Addr               { return &mockAddr{m.addr} }
+func (m *mockNetConn) SetDeadline(t time.Time) error      { return nil }
+func (m *mockNetConn) SetReadDeadline(t time.Time) error  { return nil }
+func (m *mockNetConn) SetWriteDeadline(t time.Time) error { return nil }
+
+type mockAddr struct {
+	addr string
+}
+
+func (m *mockAddr) Network() string { return "tcp" }
+func (m *mockAddr) String() string  { return m.addr }
+
+// TestEventDrivenHandoffIntegration tests the complete event-driven handoff flow
+func TestEventDrivenHandoffIntegration(t *testing.T) {
+	t.Run("EventDrivenHandoffWithPoolSkipping", func(t *testing.T) {
+		// Create a base dialer for testing
+		baseDialer := func(ctx context.Context, network, addr string) (net.Conn, error) {
+			return &mockNetConn{addr: addr}, nil
+		}
+
+		// Create processor with event-driven handoff support
+		processor := hitless.NewPoolHook(baseDialer, "tcp", nil, nil)
+		defer processor.Shutdown(context.Background())
+
+		// Create a test pool with hooks
+		hookManager := pool.NewPoolHookManager()
+		hookManager.AddHook(processor)
+
+		testPool := pool.NewConnPool(&pool.Options{
+			Dialer: func(ctx context.Context) (net.Conn, error) {
+				return &mockNetConn{addr: "original:6379"}, nil
+			},
+			PoolSize:    int32(5),
+			PoolTimeout: time.Second,
+		})
+
+		// Add the hook to the pool after creation
+		testPool.AddPoolHook(processor)
+		defer testPool.Close()
+
+		// Set the pool reference in the processor for connection removal on handoff failure
+		processor.SetPool(testPool)
+
+		ctx := context.Background()
+
+		// Get a connection and mark it for handoff
+		conn, err := testPool.Get(ctx)
+		if err != nil {
+			t.Fatalf("Failed to get connection: %v", err)
+		}
+
+		// Set initialization function with a small delay to ensure handoff is pending
+		initConnCalled := false
+		initConnFunc := func(ctx context.Context, cn *pool.Conn) error {
+			time.Sleep(50 * time.Millisecond) // Add delay to keep handoff pending
+			initConnCalled = true
+			return nil
+		}
+		conn.SetInitConnFunc(initConnFunc)
+
+		// Mark connection for handoff
+		err = conn.MarkForHandoff("new-endpoint:6379", 12345)
+		if err != nil {
+			t.Fatalf("Failed to mark connection for handoff: %v", err)
+		}
+
+		// Return connection to pool - this should queue handoff
+		testPool.Put(ctx, conn)
+
+		// Give the on-demand worker a moment to start processing
+		time.Sleep(10 * time.Millisecond)
+
+		// Verify handoff was queued
+		if !processor.IsHandoffPending(conn) {
+			t.Error("Handoff should be queued in pending map")
+		}
+
+		// Try to get the same connection - should be skipped due to pending handoff
+		conn2, err := testPool.Get(ctx)
+		if err != nil {
+			t.Fatalf("Failed to get second connection: %v", err)
+		}
+
+		// Should get a different connection (the pending one should be skipped)
+		if conn == conn2 {
+			t.Error("Should have gotten a different connection while handoff is pending")
+		}
+
+		// Return the second connection
+		testPool.Put(ctx, conn2)
+
+		// Wait for handoff to complete
+		time.Sleep(200 * time.Millisecond)
+
+		// Verify handoff completed (removed from pending map)
+		if processor.IsHandoffPending(conn) {
+			t.Error("Handoff should have completed and been removed from pending map")
+		}
+
+		if !initConnCalled {
+			t.Error("InitConn should have been called during handoff")
+		}
+
+		// Now the original connection should be available again
+		conn3, err := testPool.Get(ctx)
+		if err != nil {
+			t.Fatalf("Failed to get third connection: %v", err)
+		}
+
+		// Could be the original connection (now handed off) or a new one
+		testPool.Put(ctx, conn3)
+	})
+
+	t.Run("ConcurrentHandoffs", func(t *testing.T) {
+		// Create a base dialer that simulates slow handoffs
+		baseDialer := func(ctx context.Context, network, addr string) (net.Conn, error) {
+			time.Sleep(50 * time.Millisecond) // Simulate network delay
+			return &mockNetConn{addr: addr}, nil
+		}
+
+		processor := hitless.NewPoolHook(baseDialer, "tcp", nil, nil)
+		defer processor.Shutdown(context.Background())
+
+		// Create hooks manager and add processor as hook
+		hookManager := pool.NewPoolHookManager()
+		hookManager.AddHook(processor)
+
+		testPool := pool.NewConnPool(&pool.Options{
+			Dialer: func(ctx context.Context) (net.Conn, error) {
+				return &mockNetConn{addr: "original:6379"}, nil
+			},
+
+			PoolSize:    int32(10),
+			PoolTimeout: time.Second,
+		})
+		defer testPool.Close()
+
+		// Add the hook to the pool after creation
+		testPool.AddPoolHook(processor)
+
+		// Set the pool reference in the processor
+		processor.SetPool(testPool)
+
+		ctx := context.Background()
+		var wg sync.WaitGroup
+
+		// Start multiple concurrent handoffs
+		for i := 0; i < 5; i++ {
+			wg.Add(1)
+			go func(id int) {
+				defer wg.Done()
+
+				// Get connection
+				conn, err := testPool.Get(ctx)
+				if err != nil {
+					t.Errorf("Failed to get conn[%d]: %v", id, err)
+					return
+				}
+
+				// Set initialization function
+				initConnFunc := func(ctx context.Context, cn *pool.Conn) error {
+					return nil
+				}
+				conn.SetInitConnFunc(initConnFunc)
+
+				// Mark for handoff
+				conn.MarkForHandoff("new-endpoint:6379", int64(id))
+
+				// Return to pool (starts async handoff)
+				testPool.Put(ctx, conn)
+			}(i)
+		}
+
+		wg.Wait()
+
+		// Wait for all handoffs to complete
+		time.Sleep(300 * time.Millisecond)
+
+		// Verify pool is still functional
+		conn, err := testPool.Get(ctx)
+		if err != nil {
+			t.Fatalf("Pool should still be functional after concurrent handoffs: %v", err)
+		}
+		testPool.Put(ctx, conn)
+	})
+
+	t.Run("HandoffFailureRecovery", func(t *testing.T) {
+		// Create a failing base dialer
+		failingDialer := func(ctx context.Context, network, addr string) (net.Conn, error) {
+			return nil, &net.OpError{Op: "dial", Err: &net.DNSError{Name: addr}}
+		}
+
+		processor := hitless.NewPoolHook(failingDialer, "tcp", nil, nil)
+		defer processor.Shutdown(context.Background())
+
+		// Create hooks manager and add processor as hook
+		hookManager := pool.NewPoolHookManager()
+		hookManager.AddHook(processor)
+
+		testPool := pool.NewConnPool(&pool.Options{
+			Dialer: func(ctx context.Context) (net.Conn, error) {
+				return &mockNetConn{addr: "original:6379"}, nil
+			},
+
+			PoolSize:    int32(3),
+			PoolTimeout: time.Second,
+		})
+		defer testPool.Close()
+
+		// Add the hook to the pool after creation
+		testPool.AddPoolHook(processor)
+
+		// Set the pool reference in the processor
+		processor.SetPool(testPool)
+
+		ctx := context.Background()
+
+		// Get connection and mark for handoff
+		conn, err := testPool.Get(ctx)
+		if err != nil {
+			t.Fatalf("Failed to get connection: %v", err)
+		}
+
+		conn.MarkForHandoff("unreachable-endpoint:6379", 12345)
+
+		// Return to pool (starts async handoff that will fail)
+		testPool.Put(ctx, conn)
+
+		// Wait for handoff to fail
+		time.Sleep(200 * time.Millisecond)
+
+		// Connection should be removed from pending map after failed handoff
+		if processor.IsHandoffPending(conn) {
+			t.Error("Connection should be removed from pending map after failed handoff")
+		}
+
+		// Pool should still be functional
+		conn2, err := testPool.Get(ctx)
+		if err != nil {
+			t.Fatalf("Pool should still be functional: %v", err)
+		}
+
+		// In event-driven approach, the original connection remains in pool
+		// even after failed handoff (it's still a valid connection)
+		// We might get the same connection or a different one
+		testPool.Put(ctx, conn2)
+	})
+
+	t.Run("GracefulShutdown", func(t *testing.T) {
+		// Create a slow base dialer
+		slowDialer := func(ctx context.Context, network, addr string) (net.Conn, error) {
+			time.Sleep(100 * time.Millisecond)
+			return &mockNetConn{addr: addr}, nil
+		}
+
+		processor := hitless.NewPoolHook(slowDialer, "tcp", nil, nil)
+		defer processor.Shutdown(context.Background())
+
+		// Create hooks manager and add processor as hook
+		hookManager := pool.NewPoolHookManager()
+		hookManager.AddHook(processor)
+
+		testPool := pool.NewConnPool(&pool.Options{
+			Dialer: func(ctx context.Context) (net.Conn, error) {
+				return &mockNetConn{addr: "original:6379"}, nil
+			},
+
+			PoolSize:    int32(2),
+			PoolTimeout: time.Second,
+		})
+		defer testPool.Close()
+
+		// Add the hook to the pool after creation
+		testPool.AddPoolHook(processor)
+
+		// Set the pool reference in the processor
+		processor.SetPool(testPool)
+
+		ctx := context.Background()
+
+		// Start a handoff
+		conn, err := testPool.Get(ctx)
+		if err != nil {
+			t.Fatalf("Failed to get connection: %v", err)
+		}
+
+		if err := conn.MarkForHandoff("new-endpoint:6379", 12345); err != nil {
+			t.Fatalf("Failed to mark connection for handoff: %v", err)
+		}
+
+		// Set a mock initialization function with delay to ensure handoff is pending
+		conn.SetInitConnFunc(func(ctx context.Context, cn *pool.Conn) error {
+			time.Sleep(50 * time.Millisecond) // Add delay to keep handoff pending
+			return nil
+		})
+
+		testPool.Put(ctx, conn)
+
+		// Give the on-demand worker a moment to start and begin processing
+		// The handoff should be pending because the slowDialer takes 100ms
+		time.Sleep(10 * time.Millisecond)
+
+		// Verify handoff was queued and is being processed
+		if !processor.IsHandoffPending(conn) {
+			t.Error("Handoff should be queued in pending map")
+		}
+
+		// Give the handoff a moment to start processing
+		time.Sleep(50 * time.Millisecond)
+
+		// Shutdown processor gracefully
+		// Use a longer timeout to account for slow dialer (100ms) plus processing overhead
+		shutdownCtx, cancel := context.WithTimeout(context.Background(), 2*time.Second)
+		defer cancel()
+
+		err = processor.Shutdown(shutdownCtx)
+		if err != nil {
+			t.Errorf("Graceful shutdown should succeed: %v", err)
+		}
+
+		// Handoff should have completed (removed from pending map)
+		if processor.IsHandoffPending(conn) {
+			t.Error("Handoff should have completed and been removed from pending map after shutdown")
+		}
+	})
+}
+
+func init() {
+	logging.Disable()
+}

bench_decode_test.go

@@ -1,316 +0,0 @@
-package redis
-
-import (
-	"context"
-	"fmt"
-	"io"
-	"net"
-	"testing"
-	"time"
-
-	"github.com/redis/go-redis/v9/internal/proto"
-)
-
-var ctx = context.TODO()
-
-type ClientStub struct {
-	Cmdable
-	resp []byte
-}
-
-var initHello = []byte("%1\r\n+proto\r\n:3\r\n")
-
-func NewClientStub(resp []byte) *ClientStub {
-	stub := &ClientStub{
-		resp: resp,
-	}
-
-	stub.Cmdable = NewClient(&Options{
-		PoolSize: 128,
-		Dialer: func(ctx context.Context, network, addr string) (net.Conn, error) {
-			return stub.stubConn(initHello), nil
-		},
-		DisableIdentity: true,
-	})
-	return stub
-}
-
-func NewClusterClientStub(resp []byte) *ClientStub {
-	stub := &ClientStub{
-		resp: resp,
-	}
-
-	client := NewClusterClient(&ClusterOptions{
-		PoolSize: 128,
-		Addrs:    []string{":6379"},
-		Dialer: func(ctx context.Context, network, addr string) (net.Conn, error) {
-			return stub.stubConn(initHello), nil
-		},
-		DisableIdentity: true,
-
-		ClusterSlots: func(_ context.Context) ([]ClusterSlot, error) {
-			return []ClusterSlot{
-				{
-					Start: 0,
-					End:   16383,
-					Nodes: []ClusterNode{{Addr: "127.0.0.1:6379"}},
-				},
-			}, nil
-		},
-	})
-
-	stub.Cmdable = client
-	return stub
-}
-
-func (c *ClientStub) stubConn(init []byte) *ConnStub {
-	return &ConnStub{
-		init: init,
-		resp: c.resp,
-	}
-}
-
-type ConnStub struct {
-	init []byte
-	resp []byte
-	pos  int
-}
-
-func (c *ConnStub) Read(b []byte) (n int, err error) {
-	// Return conn.init()
-	if len(c.init) > 0 {
-		n = copy(b, c.init)
-		c.init = c.init[n:]
-		return n, nil
-	}
-
-	if len(c.resp) == 0 {
-		return 0, io.EOF
-	}
-
-	if c.pos >= len(c.resp) {
-		c.pos = 0
-	}
-	n = copy(b, c.resp[c.pos:])
-	c.pos += n
-	return n, nil
-}
-
-func (c *ConnStub) Write(b []byte) (n int, err error)  { return len(b), nil }
-func (c *ConnStub) Close() error                       { return nil }
-func (c *ConnStub) LocalAddr() net.Addr                { return nil }
-func (c *ConnStub) RemoteAddr() net.Addr               { return nil }
-func (c *ConnStub) SetDeadline(_ time.Time) error      { return nil }
-func (c *ConnStub) SetReadDeadline(_ time.Time) error  { return nil }
-func (c *ConnStub) SetWriteDeadline(_ time.Time) error { return nil }
-
-type ClientStubFunc func([]byte) *ClientStub
-
-func BenchmarkDecode(b *testing.B) {
-	type Benchmark struct {
-		name string
-		stub ClientStubFunc
-	}
-
-	benchmarks := []Benchmark{
-		{"server", NewClientStub},
-		{"cluster", NewClusterClientStub},
-	}
-
-	for _, bench := range benchmarks {
-		b.Run(fmt.Sprintf("RespError-%s", bench.name), func(b *testing.B) {
-			respError(b, bench.stub)
-		})
-		b.Run(fmt.Sprintf("RespStatus-%s", bench.name), func(b *testing.B) {
-			respStatus(b, bench.stub)
-		})
-		b.Run(fmt.Sprintf("RespInt-%s", bench.name), func(b *testing.B) {
-			respInt(b, bench.stub)
-		})
-		b.Run(fmt.Sprintf("RespString-%s", bench.name), func(b *testing.B) {
-			respString(b, bench.stub)
-		})
-		b.Run(fmt.Sprintf("RespArray-%s", bench.name), func(b *testing.B) {
-			respArray(b, bench.stub)
-		})
-		b.Run(fmt.Sprintf("RespPipeline-%s", bench.name), func(b *testing.B) {
-			respPipeline(b, bench.stub)
-		})
-		b.Run(fmt.Sprintf("RespTxPipeline-%s", bench.name), func(b *testing.B) {
-			respTxPipeline(b, bench.stub)
-		})
-
-		// goroutine
-		b.Run(fmt.Sprintf("DynamicGoroutine-%s-pool=5", bench.name), func(b *testing.B) {
-			dynamicGoroutine(b, bench.stub, 5)
-		})
-		b.Run(fmt.Sprintf("DynamicGoroutine-%s-pool=20", bench.name), func(b *testing.B) {
-			dynamicGoroutine(b, bench.stub, 20)
-		})
-		b.Run(fmt.Sprintf("DynamicGoroutine-%s-pool=50", bench.name), func(b *testing.B) {
-			dynamicGoroutine(b, bench.stub, 50)
-		})
-		b.Run(fmt.Sprintf("DynamicGoroutine-%s-pool=100", bench.name), func(b *testing.B) {
-			dynamicGoroutine(b, bench.stub, 100)
-		})
-
-		b.Run(fmt.Sprintf("StaticGoroutine-%s-pool=5", bench.name), func(b *testing.B) {
-			staticGoroutine(b, bench.stub, 5)
-		})
-		b.Run(fmt.Sprintf("StaticGoroutine-%s-pool=20", bench.name), func(b *testing.B) {
-			staticGoroutine(b, bench.stub, 20)
-		})
-		b.Run(fmt.Sprintf("StaticGoroutine-%s-pool=50", bench.name), func(b *testing.B) {
-			staticGoroutine(b, bench.stub, 50)
-		})
-		b.Run(fmt.Sprintf("StaticGoroutine-%s-pool=100", bench.name), func(b *testing.B) {
-			staticGoroutine(b, bench.stub, 100)
-		})
-	}
-}
-
-func respError(b *testing.B, stub ClientStubFunc) {
-	rdb := stub([]byte("-ERR test error\r\n"))
-	respErr := proto.RedisError("ERR test error")
-
-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
-		if err := rdb.Get(ctx, "key").Err(); err != respErr {
-			b.Fatalf("response error, got %q, want %q", err, respErr)
-		}
-	}
-}
-
-func respStatus(b *testing.B, stub ClientStubFunc) {
-	rdb := stub([]byte("+OK\r\n"))
-	var val string
-
-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
-		if val = rdb.Set(ctx, "key", "value", 0).Val(); val != "OK" {
-			b.Fatalf("response error, got %q, want OK", val)
-		}
-	}
-}
-
-func respInt(b *testing.B, stub ClientStubFunc) {
-	rdb := stub([]byte(":10\r\n"))
-	var val int64
-
-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
-		if val = rdb.Incr(ctx, "key").Val(); val != 10 {
-			b.Fatalf("response error, got %q, want 10", val)
-		}
-	}
-}
-
-func respString(b *testing.B, stub ClientStubFunc) {
-	rdb := stub([]byte("$5\r\nhello\r\n"))
-	var val string
-
-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
-		if val = rdb.Get(ctx, "key").Val(); val != "hello" {
-			b.Fatalf("response error, got %q, want hello", val)
-		}
-	}
-}
-
-func respArray(b *testing.B, stub ClientStubFunc) {
-	rdb := stub([]byte("*3\r\n$5\r\nhello\r\n:10\r\n+OK\r\n"))
-	var val []interface{}
-
-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
-		if val = rdb.MGet(ctx, "key").Val(); len(val) != 3 {
-			b.Fatalf("response error, got len(%d), want len(3)", len(val))
-		}
-	}
-}
-
-func respPipeline(b *testing.B, stub ClientStubFunc) {
-	rdb := stub([]byte("+OK\r\n$5\r\nhello\r\n:1\r\n"))
-	var pipe Pipeliner
-
-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
-		pipe = rdb.Pipeline()
-		set := pipe.Set(ctx, "key", "value", 0)
-		get := pipe.Get(ctx, "key")
-		del := pipe.Del(ctx, "key")
-		_, err := pipe.Exec(ctx)
-		if err != nil {
-			b.Fatalf("response error, got %q, want nil", err)
-		}
-		if set.Val() != "OK" || get.Val() != "hello" || del.Val() != 1 {
-			b.Fatal("response error")
-		}
-	}
-}
-
-func respTxPipeline(b *testing.B, stub ClientStubFunc) {
-	rdb := stub([]byte("+OK\r\n+QUEUED\r\n+QUEUED\r\n+QUEUED\r\n*3\r\n+OK\r\n$5\r\nhello\r\n:1\r\n"))
-
-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
-		var set *StatusCmd
-		var get *StringCmd
-		var del *IntCmd
-		_, err := rdb.TxPipelined(ctx, func(pipe Pipeliner) error {
-			set = pipe.Set(ctx, "key", "value", 0)
-			get = pipe.Get(ctx, "key")
-			del = pipe.Del(ctx, "key")
-			return nil
-		})
-		if err != nil {
-			b.Fatalf("response error, got %q, want nil", err)
-		}
-		if set.Val() != "OK" || get.Val() != "hello" || del.Val() != 1 {
-			b.Fatal("response error")
-		}
-	}
-}
-
-func dynamicGoroutine(b *testing.B, stub ClientStubFunc, concurrency int) {
-	rdb := stub([]byte("$5\r\nhello\r\n"))
-	c := make(chan struct{}, concurrency)
-
-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
-		c <- struct{}{}
-		go func() {
-			if val := rdb.Get(ctx, "key").Val(); val != "hello" {
-				panic(fmt.Sprintf("response error, got %q, want hello", val))
-			}
-			<-c
-		}()
-	}
-	// Here no longer wait for all goroutines to complete, it will not affect the test results.
-	close(c)
-}
-
-func staticGoroutine(b *testing.B, stub ClientStubFunc, concurrency int) {
-	rdb := stub([]byte("$5\r\nhello\r\n"))
-	c := make(chan struct{}, concurrency)
-
-	b.ResetTimer()
-
-	for i := 0; i < concurrency; i++ {
-		go func() {
-			for {
-				_, ok := <-c
-				if !ok {
-					return
-				}
-				if val := rdb.Get(ctx, "key").Val(); val != "hello" {
-					panic(fmt.Sprintf("response error, got %q, want hello", val))
-				}
-			}
-		}()
-	}
-	for i := 0; i < b.N; i++ {
-		c <- struct{}{}
-	}
-	close(c)
-}

commands.go

@@ -193,6 +193,7 @@ type Cmdable interface {
 	ClientID(ctx context.Context) *IntCmd
 	ClientUnblock(ctx context.Context, id int64) *IntCmd
 	ClientUnblockWithError(ctx context.Context, id int64) *IntCmd
+	ClientMaintNotifications(ctx context.Context, enabled bool, endpointType string) *StatusCmd
 	ConfigGet(ctx context.Context, parameter string) *MapStringStringCmd
 	ConfigResetStat(ctx context.Context) *StatusCmd
 	ConfigSet(ctx context.Context, parameter, value string) *StatusCmd
@@ -519,6 +520,23 @@ func (c cmdable) ClientInfo(ctx context.Context) *ClientInfoCmd {
 	return cmd
 }
 
+// ClientMaintNotifications enables or disables maintenance notifications for hitless upgrades.
+// When enabled, the client will receive push notifications about Redis maintenance events.
+func (c cmdable) ClientMaintNotifications(ctx context.Context, enabled bool, endpointType string) *StatusCmd {
+	args := []interface{}{"client", "maint_notifications"}
+	if enabled {
+		if endpointType == "" {
+			endpointType = "none"
+		}
+		args = append(args, "on", "moving-endpoint-type", endpointType)
+	} else {
+		args = append(args, "off")
+	}
+	cmd := NewStatusCmd(ctx, args...)
+	_ = c(ctx, cmd)
+	return cmd
+}
+
 // ------------------------------------------------------------------------------------------------
 
 func (c cmdable) ConfigGet(ctx context.Context, parameter string) *MapStringStringCmd {

commands_test.go

@@ -3019,7 +3019,8 @@ var _ = Describe("Commands", func() {
 
 			res, err = client.HPTTL(ctx, "myhash", "key1", "key2", "key200").Result()
 			Expect(err).NotTo(HaveOccurred())
-			Expect(res[0]).To(BeNumerically("~", 10*time.Second.Milliseconds(), 1))
+			// overhead of the push notification check is about 1-2ms for 100 commands
+			Expect(res[0]).To(BeNumerically("~", 10*time.Second.Milliseconds(), 2))
 		})
 
 		It("should HGETDEL", Label("hash", "HGETDEL"), func() {

example/pubsub/go.mod

@@ -0,0 +1,12 @@
+module github.com/redis/go-redis/example/pubsub
+
+go 1.18
+
+replace github.com/redis/go-redis/v9 => ../..
+
+require github.com/redis/go-redis/v9 v9.11.0
+
+require (
+	github.com/cespare/xxhash/v2 v2.3.0 // indirect
+	github.com/dgryski/go-rendezvous v0.0.0-20200823014737-9f7001d12a5f // indirect
+)

example/pubsub/go.sum

@@ -0,0 +1,6 @@
+github.com/bsm/ginkgo/v2 v2.12.0 h1:Ny8MWAHyOepLGlLKYmXG4IEkioBysk6GpaRTLC8zwWs=
+github.com/bsm/gomega v1.27.10 h1:yeMWxP2pV2fG3FgAODIY8EiRE3dy0aeFYt4l7wh6yKA=
+github.com/cespare/xxhash/v2 v2.3.0 h1:UL815xU9SqsFlibzuggzjXhog7bL6oX9BbNZnL2UFvs=
+github.com/cespare/xxhash/v2 v2.3.0/go.mod h1:VGX0DQ3Q6kWi7AoAeZDth3/j3BFtOZR5XLFGgcrjCOs=
+github.com/dgryski/go-rendezvous v0.0.0-20200823014737-9f7001d12a5f h1:lO4WD4F/rVNCu3HqELle0jiPLLBs70cWOduZpkS1E78=
+github.com/dgryski/go-rendezvous v0.0.0-20200823014737-9f7001d12a5f/go.mod h1:cuUVRXasLTGF7a8hSLbxyZXjz+1KgoB3wDUb6vlszIc=

example/pubsub/main.go

@@ -0,0 +1,171 @@
+package main
+
+import (
+	"context"
+	"fmt"
+	"log"
+	"sync"
+	"sync/atomic"
+	"time"
+
+	"github.com/redis/go-redis/v9"
+	"github.com/redis/go-redis/v9/hitless"
+	"github.com/redis/go-redis/v9/logging"
+)
+
+var ctx = context.Background()
+var cntErrors atomic.Int64
+var cntSuccess atomic.Int64
+var startTime = time.Now()
+
+// This example is not supposed to be run as is. It is just a test to see how pubsub behaves in relation to pool management.
+// It was used to find regressions in pool management in hitless mode.
+// Please don't use it as a reference for how to use pubsub.
+func main() {
+	startTime = time.Now()
+	wg := &sync.WaitGroup{}
+	rdb := redis.NewClient(&redis.Options{
+		Addr: ":6379",
+		HitlessUpgradeConfig: &redis.HitlessUpgradeConfig{
+			Mode: hitless.MaintNotificationsEnabled,
+		},
+	})
+	_ = rdb.FlushDB(ctx).Err()
+	hitlessManager := rdb.GetHitlessManager()
+	if hitlessManager == nil {
+		panic("hitless manager is nil")
+	}
+	loggingHook := hitless.NewLoggingHook(logging.LogLevelDebug)
+	hitlessManager.AddNotificationHook(loggingHook)
+
+	go func() {
+		for {
+			time.Sleep(2 * time.Second)
+			fmt.Printf("pool stats: %+v\n", rdb.PoolStats())
+		}
+	}()
+	err := rdb.Ping(ctx).Err()
+	if err != nil {
+		panic(err)
+	}
+	if err := rdb.Set(ctx, "publishers", "0", 0).Err(); err != nil {
+		panic(err)
+	}
+	if err := rdb.Set(ctx, "subscribers", "0", 0).Err(); err != nil {
+		panic(err)
+	}
+	if err := rdb.Set(ctx, "published", "0", 0).Err(); err != nil {
+		panic(err)
+	}
+	if err := rdb.Set(ctx, "received", "0", 0).Err(); err != nil {
+		panic(err)
+	}
+	fmt.Println("published", rdb.Get(ctx, "published").Val())
+	fmt.Println("received", rdb.Get(ctx, "received").Val())
+	subCtx, cancelSubCtx := context.WithCancel(ctx)
+	pubCtx, cancelPublishers := context.WithCancel(ctx)
+	for i := 0; i < 10; i++ {
+		wg.Add(1)
+		go subscribe(subCtx, rdb, "test", i, wg)
+	}
+	time.Sleep(time.Second)
+	cancelSubCtx()
+	time.Sleep(time.Second)
+	subCtx, cancelSubCtx = context.WithCancel(ctx)
+	for i := 0; i < 10; i++ {
+		if err := rdb.Incr(ctx, "publishers").Err(); err != nil {
+			fmt.Println("incr error:", err)
+			cntErrors.Add(1)
+		}
+		wg.Add(1)
+		go floodThePool(pubCtx, rdb, wg)
+	}
+
+	for i := 0; i < 500; i++ {
+		if err := rdb.Incr(ctx, "subscribers").Err(); err != nil {
+			fmt.Println("incr error:", err)
+			cntErrors.Add(1)
+		}
+
+		wg.Add(1)
+		go subscribe(subCtx, rdb, "test2", i, wg)
+	}
+	time.Sleep(120 * time.Second)
+	fmt.Println("canceling publishers")
+	cancelPublishers()
+	time.Sleep(10 * time.Second)
+	fmt.Println("canceling subscribers")
+	cancelSubCtx()
+	wg.Wait()
+	published, err := rdb.Get(ctx, "published").Result()
+	received, err := rdb.Get(ctx, "received").Result()
+	publishers, err := rdb.Get(ctx, "publishers").Result()
+	subscribers, err := rdb.Get(ctx, "subscribers").Result()
+	fmt.Printf("publishers: %s\n", publishers)
+	fmt.Printf("published: %s\n", published)
+	fmt.Printf("subscribers: %s\n", subscribers)
+	fmt.Printf("received: %s\n", received)
+	publishedInt, err := rdb.Get(ctx, "published").Int()
+	subscribersInt, err := rdb.Get(ctx, "subscribers").Int()
+	fmt.Printf("if drained = published*subscribers: %d\n", publishedInt*subscribersInt)
+
+	time.Sleep(2 * time.Second)
+	fmt.Println("errors:", cntErrors.Load())
+	fmt.Println("success:", cntSuccess.Load())
+	fmt.Println("time:", time.Since(startTime))
+}
+
+func floodThePool(ctx context.Context, rdb *redis.Client, wg *sync.WaitGroup) {
+	defer wg.Done()
+	for {
+		select {
+		case <-ctx.Done():
+			return
+		default:
+		}
+		err := rdb.Publish(ctx, "test2", "hello").Err()
+		if err != nil {
+			if err.Error() != "context canceled" {
+				log.Println("publish error:", err)
+				cntErrors.Add(1)
+			}
+		}
+
+		err = rdb.Incr(ctx, "published").Err()
+		if err != nil {
+			if err.Error() != "context canceled" {
+				log.Println("incr error:", err)
+				cntErrors.Add(1)
+			}
+		}
+		time.Sleep(10 * time.Nanosecond)
+	}
+}
+
+func subscribe(ctx context.Context, rdb *redis.Client, topic string, subscriberId int, wg *sync.WaitGroup) {
+	defer wg.Done()
+	rec := rdb.Subscribe(ctx, topic)
+	recChan := rec.Channel()
+	for {
+		select {
+		case <-ctx.Done():
+			rec.Close()
+			return
+		default:
+			select {
+			case <-ctx.Done():
+				rec.Close()
+				return
+			case msg := <-recChan:
+				err := rdb.Incr(ctx, "received").Err()
+				if err != nil {
+					if err.Error() != "context canceled" {
+						log.Printf("%s\n", err.Error())
+						cntErrors.Add(1)
+					}
+				}
+				_ = msg // Use the message to avoid unused variable warning
+			}
+		}
+	}
+}

example_instrumentation_test.go

@@ -57,6 +57,8 @@ func Example_instrumentation() {
 	// finished dialing tcp :6379
 	// starting processing: <[hello 3]>
 	// finished processing: <[hello 3]>
+	// starting processing: <[client maint_notifications on moving-endpoint-type internal-fqdn]>
+	// finished processing: <[client maint_notifications on moving-endpoint-type internal-fqdn]>
 	// finished processing: <[ping]>
 }
 
@@ -78,6 +80,8 @@ func ExamplePipeline_instrumentation() {
 	// finished dialing tcp :6379
 	// starting processing: <[hello 3]>
 	// finished processing: <[hello 3]>
+	// starting processing: <[client maint_notifications on moving-endpoint-type internal-fqdn]>
+	// finished processing: <[client maint_notifications on moving-endpoint-type internal-fqdn]>
 	// pipeline finished processing: [[ping] [ping]]
 }
 
@@ -99,6 +103,8 @@ func ExampleClient_Watch_instrumentation() {
 	// finished dialing tcp :6379
 	// starting processing: <[hello 3]>
 	// finished processing: <[hello 3]>
+	// starting processing: <[client maint_notifications on moving-endpoint-type internal-fqdn]>
+	// finished processing: <[client maint_notifications on moving-endpoint-type internal-fqdn]>
 	// finished processing: <[watch foo]>
 	// starting processing: <[ping]>
 	// finished processing: <[ping]>

hitless/README.md

@@ -0,0 +1,98 @@
+# Hitless Upgrades
+
+Seamless Redis connection handoffs during cluster changes without dropping connections.
+
+## Quick Start
+
+```go
+client := redis.NewClient(&redis.Options{
+    Addr:     "localhost:6379",
+    Protocol: 3, // RESP3 required
+    HitlessUpgrades: &hitless.Config{
+        Mode: hitless.MaintNotificationsEnabled,
+    },
+})
+```
+
+## Modes
+
+- **`MaintNotificationsDisabled`** - Hitless upgrades disabled
+- **`MaintNotificationsEnabled`** - Forcefully enabled (fails if server doesn't support)
+- **`MaintNotificationsAuto`** - Auto-detect server support (default)
+
+## Configuration
+
+```go
+&hitless.Config{
+    Mode:                       hitless.MaintNotificationsAuto,
+    EndpointType:               hitless.EndpointTypeAuto,
+    RelaxedTimeout:             10 * time.Second,
+    HandoffTimeout:             15 * time.Second,
+    MaxHandoffRetries:          3,
+    MaxWorkers:                 0,    // Auto-calculated
+    HandoffQueueSize:           0,    // Auto-calculated
+    PostHandoffRelaxedDuration: 0,    // 2 * RelaxedTimeout
+    LogLevel:                   logging.LogLevelError,
+}
+```
+
+### Endpoint Types
+
+- **`EndpointTypeAuto`** - Auto-detect based on connection (default)
+- **`EndpointTypeInternalIP`** - Internal IP address
+- **`EndpointTypeInternalFQDN`** - Internal FQDN
+- **`EndpointTypeExternalIP`** - External IP address
+- **`EndpointTypeExternalFQDN`** - External FQDN
+- **`EndpointTypeNone`** - No endpoint (reconnect with current config)
+
+### Auto-Scaling
+
+**Workers**: `min(PoolSize/2, max(10, PoolSize/3))` when auto-calculated
+**Queue**: `max(20×Workers, PoolSize)` capped by `MaxActiveConns+1` or `5×PoolSize`
+
+**Examples:**
+- Pool 100: 33 workers, 660 queue (capped at 500)
+- Pool 100 + MaxActiveConns 150: 33 workers, 151 queue
+
+## How It Works
+
+1. Redis sends push notifications about cluster changes
+2. Client creates new connections to updated endpoints
+3. Active operations transfer to new connections
+4. Old connections close gracefully
+
+## Supported Notifications
+
+- `MOVING` - Slot moving to new node
+- `MIGRATING` - Slot in migration state
+- `MIGRATED` - Migration completed
+- `FAILING_OVER` - Node failing over
+- `FAILED_OVER` - Failover completed
+
+## Hooks (Optional)
+
+Monitor and customize hitless operations:
+
+```go
+type NotificationHook interface {
+    PreHook(ctx, notificationCtx, notificationType, notification) ([]interface{}, bool)
+    PostHook(ctx, notificationCtx, notificationType, notification, result)
+}
+
+// Add custom hook
+manager.AddNotificationHook(&MyHook{})
+```
+
+### Metrics Hook Example
+
+```go
+// Create metrics hook
+metricsHook := hitless.NewMetricsHook()
+manager.AddNotificationHook(metricsHook)
+
+// Access collected metrics
+metrics := metricsHook.GetMetrics()
+fmt.Printf("Notification counts: %v\n", metrics["notification_counts"])
+fmt.Printf("Processing times: %v\n", metrics["processing_times"])
+fmt.Printf("Error counts: %v\n", metrics["error_counts"])
+```

hitless/circuit_breaker.go

@@ -0,0 +1,360 @@
+package hitless
+
+import (
+	"context"
+	"sync"
+	"sync/atomic"
+	"time"
+
+	"github.com/redis/go-redis/v9/internal"
+)
+
+// CircuitBreakerState represents the state of a circuit breaker
+type CircuitBreakerState int32
+
+const (
+	// CircuitBreakerClosed - normal operation, requests allowed
+	CircuitBreakerClosed CircuitBreakerState = iota
+	// CircuitBreakerOpen - failing fast, requests rejected
+	CircuitBreakerOpen
+	// CircuitBreakerHalfOpen - testing if service recovered
+	CircuitBreakerHalfOpen
+)
+
+func (s CircuitBreakerState) String() string {
+	switch s {
+	case CircuitBreakerClosed:
+		return "closed"
+	case CircuitBreakerOpen:
+		return "open"
+	case CircuitBreakerHalfOpen:
+		return "half-open"
+	default:
+		return "unknown"
+	}
+}
+
+// CircuitBreaker implements the circuit breaker pattern for endpoint-specific failure handling
+type CircuitBreaker struct {
+	// Configuration
+	failureThreshold int           // Number of failures before opening
+	resetTimeout     time.Duration // How long to stay open before testing
+	maxRequests      int           // Max requests allowed in half-open state
+
+	// State tracking (atomic for lock-free access)
+	state           atomic.Int32 // CircuitBreakerState
+	failures        atomic.Int64 // Current failure count
+	successes       atomic.Int64 // Success count in half-open state
+	requests        atomic.Int64 // Request count in half-open state
+	lastFailureTime atomic.Int64 // Unix timestamp of last failure
+	lastSuccessTime atomic.Int64 // Unix timestamp of last success
+
+	// Endpoint identification
+	endpoint string
+	config   *Config
+}
+
+// newCircuitBreaker creates a new circuit breaker for an endpoint
+func newCircuitBreaker(endpoint string, config *Config) *CircuitBreaker {
+	// Use configuration values with sensible defaults
+	failureThreshold := 5
+	resetTimeout := 60 * time.Second
+	maxRequests := 3
+
+	if config != nil {
+		failureThreshold = config.CircuitBreakerFailureThreshold
+		resetTimeout = config.CircuitBreakerResetTimeout
+		maxRequests = config.CircuitBreakerMaxRequests
+	}
+
+	return &CircuitBreaker{
+		failureThreshold: failureThreshold,
+		resetTimeout:     resetTimeout,
+		maxRequests:      maxRequests,
+		endpoint:         endpoint,
+		config:           config,
+		state:            atomic.Int32{}, // Defaults to CircuitBreakerClosed (0)
+	}
+}
+
+// IsOpen returns true if the circuit breaker is open (rejecting requests)
+func (cb *CircuitBreaker) IsOpen() bool {
+	state := CircuitBreakerState(cb.state.Load())
+	return state == CircuitBreakerOpen
+}
+
+// shouldAttemptReset checks if enough time has passed to attempt reset
+func (cb *CircuitBreaker) shouldAttemptReset() bool {
+	lastFailure := time.Unix(cb.lastFailureTime.Load(), 0)
+	return time.Since(lastFailure) >= cb.resetTimeout
+}
+
+// Execute runs the given function with circuit breaker protection
+func (cb *CircuitBreaker) Execute(fn func() error) error {
+	// Single atomic state load for consistency
+	state := CircuitBreakerState(cb.state.Load())
+
+	switch state {
+	case CircuitBreakerOpen:
+		if cb.shouldAttemptReset() {
+			// Attempt transition to half-open
+			if cb.state.CompareAndSwap(int32(CircuitBreakerOpen), int32(CircuitBreakerHalfOpen)) {
+				cb.requests.Store(0)
+				cb.successes.Store(0)
+				if cb.config != nil && cb.config.LogLevel.InfoOrAbove() {
+					internal.Logger.Printf(context.Background(),
+						"hitless: circuit breaker for %s transitioning to half-open", cb.endpoint)
+				}
+				// Fall through to half-open logic
+			} else {
+				return ErrCircuitBreakerOpen
+			}
+		} else {
+			return ErrCircuitBreakerOpen
+		}
+		fallthrough
+	case CircuitBreakerHalfOpen:
+		requests := cb.requests.Add(1)
+		if requests > int64(cb.maxRequests) {
+			cb.requests.Add(-1) // Revert the increment
+			return ErrCircuitBreakerOpen
+		}
+	}
+
+	// Execute the function with consistent state
+	err := fn()
+
+	if err != nil {
+		cb.recordFailure()
+		return err
+	}
+
+	cb.recordSuccess()
+	return nil
+}
+
+// recordFailure records a failure and potentially opens the circuit
+func (cb *CircuitBreaker) recordFailure() {
+	cb.lastFailureTime.Store(time.Now().Unix())
+	failures := cb.failures.Add(1)
+
+	state := CircuitBreakerState(cb.state.Load())
+
+	switch state {
+	case CircuitBreakerClosed:
+		if failures >= int64(cb.failureThreshold) {
+			if cb.state.CompareAndSwap(int32(CircuitBreakerClosed), int32(CircuitBreakerOpen)) {
+				if cb.config != nil && cb.config.LogLevel.WarnOrAbove() {
+					internal.Logger.Printf(context.Background(),
+						"hitless: circuit breaker opened for endpoint %s after %d failures",
+						cb.endpoint, failures)
+				}
+			}
+		}
+	case CircuitBreakerHalfOpen:
+		// Any failure in half-open state immediately opens the circuit
+		if cb.state.CompareAndSwap(int32(CircuitBreakerHalfOpen), int32(CircuitBreakerOpen)) {
+			if cb.config != nil && cb.config.LogLevel.WarnOrAbove() {
+				internal.Logger.Printf(context.Background(),
+					"hitless: circuit breaker reopened for endpoint %s due to failure in half-open state",
+					cb.endpoint)
+			}
+		}
+	}
+}
+
+// recordSuccess records a success and potentially closes the circuit
+func (cb *CircuitBreaker) recordSuccess() {
+	cb.lastSuccessTime.Store(time.Now().Unix())
+
+	state := CircuitBreakerState(cb.state.Load())
+
+	switch state {
+	case CircuitBreakerClosed:
+		// Reset failure count on success in closed state
+		cb.failures.Store(0)
+	case CircuitBreakerHalfOpen:
+		successes := cb.successes.Add(1)
+
+		// If we've had enough successful requests, close the circuit
+		if successes >= int64(cb.maxRequests) {
+			if cb.state.CompareAndSwap(int32(CircuitBreakerHalfOpen), int32(CircuitBreakerClosed)) {
+				cb.failures.Store(0)
+				if cb.config != nil && cb.config.LogLevel.InfoOrAbove() {
+					internal.Logger.Printf(context.Background(),
+						"hitless: circuit breaker closed for endpoint %s after %d successful requests",
+						cb.endpoint, successes)
+				}
+			}
+		}
+	}
+}
+
+// GetState returns the current state of the circuit breaker
+func (cb *CircuitBreaker) GetState() CircuitBreakerState {
+	return CircuitBreakerState(cb.state.Load())
+}
+
+// GetStats returns current statistics for monitoring
+func (cb *CircuitBreaker) GetStats() CircuitBreakerStats {
+	return CircuitBreakerStats{
+		Endpoint:        cb.endpoint,
+		State:           cb.GetState(),
+		Failures:        cb.failures.Load(),
+		Successes:       cb.successes.Load(),
+		Requests:        cb.requests.Load(),
+		LastFailureTime: time.Unix(cb.lastFailureTime.Load(), 0),
+		LastSuccessTime: time.Unix(cb.lastSuccessTime.Load(), 0),
+	}
+}
+
+// CircuitBreakerStats provides statistics about a circuit breaker
+type CircuitBreakerStats struct {
+	Endpoint        string
+	State           CircuitBreakerState
+	Failures        int64
+	Successes       int64
+	Requests        int64
+	LastFailureTime time.Time
+	LastSuccessTime time.Time
+}
+
+// CircuitBreakerEntry wraps a circuit breaker with access tracking
+type CircuitBreakerEntry struct {
+	breaker    *CircuitBreaker
+	lastAccess atomic.Int64 // Unix timestamp
+	created    time.Time
+}
+
+// CircuitBreakerManager manages circuit breakers for multiple endpoints
+type CircuitBreakerManager struct {
+	breakers    sync.Map // map[string]*CircuitBreakerEntry
+	config      *Config
+	cleanupStop chan struct{}
+	cleanupMu   sync.Mutex
+	lastCleanup atomic.Int64 // Unix timestamp
+}
+
+// newCircuitBreakerManager creates a new circuit breaker manager
+func newCircuitBreakerManager(config *Config) *CircuitBreakerManager {
+	cbm := &CircuitBreakerManager{
+		config:      config,
+		cleanupStop: make(chan struct{}),
+	}
+	cbm.lastCleanup.Store(time.Now().Unix())
+
+	// Start background cleanup goroutine
+	go cbm.cleanupLoop()
+
+	return cbm
+}
+
+// GetCircuitBreaker returns the circuit breaker for an endpoint, creating it if necessary
+func (cbm *CircuitBreakerManager) GetCircuitBreaker(endpoint string) *CircuitBreaker {
+	now := time.Now().Unix()
+
+	if entry, ok := cbm.breakers.Load(endpoint); ok {
+		cbEntry := entry.(*CircuitBreakerEntry)
+		cbEntry.lastAccess.Store(now)
+		return cbEntry.breaker
+	}
+
+	// Create new circuit breaker with metadata
+	newBreaker := newCircuitBreaker(endpoint, cbm.config)
+	newEntry := &CircuitBreakerEntry{
+		breaker: newBreaker,
+		created: time.Now(),
+	}
+	newEntry.lastAccess.Store(now)
+
+	actual, _ := cbm.breakers.LoadOrStore(endpoint, newEntry)
+	return actual.(*CircuitBreakerEntry).breaker
+}
+
+// GetAllStats returns statistics for all circuit breakers
+func (cbm *CircuitBreakerManager) GetAllStats() []CircuitBreakerStats {
+	var stats []CircuitBreakerStats
+	cbm.breakers.Range(func(key, value interface{}) bool {
+		entry := value.(*CircuitBreakerEntry)
+		stats = append(stats, entry.breaker.GetStats())
+		return true
+	})
+	return stats
+}
+
+// cleanupLoop runs background cleanup of unused circuit breakers
+func (cbm *CircuitBreakerManager) cleanupLoop() {
+	ticker := time.NewTicker(5 * time.Minute) // Cleanup every 5 minutes
+	defer ticker.Stop()
+
+	for {
+		select {
+		case <-ticker.C:
+			cbm.cleanup()
+		case <-cbm.cleanupStop:
+			return
+		}
+	}
+}
+
+// cleanup removes circuit breakers that haven't been accessed recently
+func (cbm *CircuitBreakerManager) cleanup() {
+	// Prevent concurrent cleanups
+	if !cbm.cleanupMu.TryLock() {
+		return
+	}
+	defer cbm.cleanupMu.Unlock()
+
+	now := time.Now()
+	cutoff := now.Add(-30 * time.Minute).Unix() // 30 minute TTL
+
+	var toDelete []string
+	count := 0
+
+	cbm.breakers.Range(func(key, value interface{}) bool {
+		endpoint := key.(string)
+		entry := value.(*CircuitBreakerEntry)
+
+		count++
+
+		// Remove if not accessed recently
+		if entry.lastAccess.Load() < cutoff {
+			toDelete = append(toDelete, endpoint)
+		}
+
+		return true
+	})
+
+	// Delete expired entries
+	for _, endpoint := range toDelete {
+		cbm.breakers.Delete(endpoint)
+	}
+
+	// Log cleanup results
+	if len(toDelete) > 0 && cbm.config != nil && cbm.config.LogLevel.InfoOrAbove() {
+		internal.Logger.Printf(context.Background(),
+			"hitless: circuit breaker cleanup removed %d/%d entries", len(toDelete), count)
+	}
+
+	cbm.lastCleanup.Store(now.Unix())
+}
+
+// Shutdown stops the cleanup goroutine
+func (cbm *CircuitBreakerManager) Shutdown() {
+	close(cbm.cleanupStop)
+}
+
+// Reset resets all circuit breakers (useful for testing)
+func (cbm *CircuitBreakerManager) Reset() {
+	cbm.breakers.Range(func(key, value interface{}) bool {
+		entry := value.(*CircuitBreakerEntry)
+		breaker := entry.breaker
+		breaker.state.Store(int32(CircuitBreakerClosed))
+		breaker.failures.Store(0)
+		breaker.successes.Store(0)
+		breaker.requests.Store(0)
+		breaker.lastFailureTime.Store(0)
+		breaker.lastSuccessTime.Store(0)
+		return true
+	})
+}

hitless/circuit_breaker_test.go

@@ -0,0 +1,356 @@
+package hitless
+
+import (
+	"errors"
+	"testing"
+	"time"
+
+	"github.com/redis/go-redis/v9/logging"
+)
+
+func TestCircuitBreaker(t *testing.T) {
+	config := &Config{
+		LogLevel:                       logging.LogLevelError, // Reduce noise in tests
+		CircuitBreakerFailureThreshold: 5,
+		CircuitBreakerResetTimeout:     60 * time.Second,
+		CircuitBreakerMaxRequests:      3,
+	}
+
+	t.Run("InitialState", func(t *testing.T) {
+		cb := newCircuitBreaker("test-endpoint:6379", config)
+
+		if cb.IsOpen() {
+			t.Error("Circuit breaker should start in closed state")
+		}
+
+		if cb.GetState() != CircuitBreakerClosed {
+			t.Errorf("Expected state %v, got %v", CircuitBreakerClosed, cb.GetState())
+		}
+	})
+
+	t.Run("SuccessfulExecution", func(t *testing.T) {
+		cb := newCircuitBreaker("test-endpoint:6379", config)
+
+		err := cb.Execute(func() error {
+			return nil // Success
+		})
+
+		if err != nil {
+			t.Errorf("Expected no error, got %v", err)
+		}
+
+		if cb.GetState() != CircuitBreakerClosed {
+			t.Errorf("Expected state %v, got %v", CircuitBreakerClosed, cb.GetState())
+		}
+	})
+
+	t.Run("FailureThreshold", func(t *testing.T) {
+		cb := newCircuitBreaker("test-endpoint:6379", config)
+		testError := errors.New("test error")
+
+		// Fail 4 times (below threshold of 5)
+		for i := 0; i < 4; i++ {
+			err := cb.Execute(func() error {
+				return testError
+			})
+			if err != testError {
+				t.Errorf("Expected test error, got %v", err)
+			}
+			if cb.GetState() != CircuitBreakerClosed {
+				t.Errorf("Circuit should still be closed after %d failures", i+1)
+			}
+		}
+
+		// 5th failure should open the circuit
+		err := cb.Execute(func() error {
+			return testError
+		})
+		if err != testError {
+			t.Errorf("Expected test error, got %v", err)
+		}
+
+		if cb.GetState() != CircuitBreakerOpen {
+			t.Errorf("Expected state %v, got %v", CircuitBreakerOpen, cb.GetState())
+		}
+	})
+
+	t.Run("OpenCircuitFailsFast", func(t *testing.T) {
+		cb := newCircuitBreaker("test-endpoint:6379", config)
+		testError := errors.New("test error")
+
+		// Force circuit to open
+		for i := 0; i < 5; i++ {
+			cb.Execute(func() error { return testError })
+		}
+
+		// Now it should fail fast
+		err := cb.Execute(func() error {
+			t.Error("Function should not be called when circuit is open")
+			return nil
+		})
+
+		if err != ErrCircuitBreakerOpen {
+			t.Errorf("Expected ErrCircuitBreakerOpen, got %v", err)
+		}
+	})
+
+	t.Run("HalfOpenTransition", func(t *testing.T) {
+		testConfig := &Config{
+			LogLevel:                       logging.LogLevelError,
+			CircuitBreakerFailureThreshold: 5,
+			CircuitBreakerResetTimeout:     100 * time.Millisecond, // Short timeout for testing
+			CircuitBreakerMaxRequests:      3,
+		}
+		cb := newCircuitBreaker("test-endpoint:6379", testConfig)
+		testError := errors.New("test error")
+
+		// Force circuit to open
+		for i := 0; i < 5; i++ {
+			cb.Execute(func() error { return testError })
+		}
+
+		if cb.GetState() != CircuitBreakerOpen {
+			t.Error("Circuit should be open")
+		}
+
+		// Wait for reset timeout
+		time.Sleep(150 * time.Millisecond)
+
+		// Next call should transition to half-open
+		executed := false
+		err := cb.Execute(func() error {
+			executed = true
+			return nil // Success
+		})
+
+		if err != nil {
+			t.Errorf("Expected no error, got %v", err)
+		}
+
+		if !executed {
+			t.Error("Function should have been executed in half-open state")
+		}
+	})
+
+	t.Run("HalfOpenToClosedTransition", func(t *testing.T) {
+		testConfig := &Config{
+			LogLevel:                       logging.LogLevelError,
+			CircuitBreakerFailureThreshold: 5,
+			CircuitBreakerResetTimeout:     50 * time.Millisecond,
+			CircuitBreakerMaxRequests:      3,
+		}
+		cb := newCircuitBreaker("test-endpoint:6379", testConfig)
+		testError := errors.New("test error")
+
+		// Force circuit to open
+		for i := 0; i < 5; i++ {
+			cb.Execute(func() error { return testError })
+		}
+
+		// Wait for reset timeout
+		time.Sleep(100 * time.Millisecond)
+
+		// Execute successful requests in half-open state
+		for i := 0; i < 3; i++ {
+			err := cb.Execute(func() error {
+				return nil // Success
+			})
+			if err != nil {
+				t.Errorf("Expected no error on attempt %d, got %v", i+1, err)
+			}
+		}
+
+		// Circuit should now be closed
+		if cb.GetState() != CircuitBreakerClosed {
+			t.Errorf("Expected state %v, got %v", CircuitBreakerClosed, cb.GetState())
+		}
+	})
+
+	t.Run("HalfOpenToOpenOnFailure", func(t *testing.T) {
+		testConfig := &Config{
+			LogLevel:                       logging.LogLevelError,
+			CircuitBreakerFailureThreshold: 5,
+			CircuitBreakerResetTimeout:     50 * time.Millisecond,
+			CircuitBreakerMaxRequests:      3,
+		}
+		cb := newCircuitBreaker("test-endpoint:6379", testConfig)
+		testError := errors.New("test error")
+
+		// Force circuit to open
+		for i := 0; i < 5; i++ {
+			cb.Execute(func() error { return testError })
+		}
+
+		// Wait for reset timeout
+		time.Sleep(100 * time.Millisecond)
+
+		// First request in half-open state fails
+		err := cb.Execute(func() error {
+			return testError
+		})
+
+		if err != testError {
+			t.Errorf("Expected test error, got %v", err)
+		}
+
+		// Circuit should be open again
+		if cb.GetState() != CircuitBreakerOpen {
+			t.Errorf("Expected state %v, got %v", CircuitBreakerOpen, cb.GetState())
+		}
+	})
+
+	t.Run("Stats", func(t *testing.T) {
+		cb := newCircuitBreaker("test-endpoint:6379", config)
+		testError := errors.New("test error")
+
+		// Execute some operations
+		cb.Execute(func() error { return testError }) // Failure
+		cb.Execute(func() error { return testError }) // Failure
+
+		stats := cb.GetStats()
+
+		if stats.Endpoint != "test-endpoint:6379" {
+			t.Errorf("Expected endpoint 'test-endpoint:6379', got %s", stats.Endpoint)
+		}
+
+		if stats.Failures != 2 {
+			t.Errorf("Expected 2 failures, got %d", stats.Failures)
+		}
+
+		if stats.State != CircuitBreakerClosed {
+			t.Errorf("Expected state %v, got %v", CircuitBreakerClosed, stats.State)
+		}
+
+		// Test that success resets failure count
+		cb.Execute(func() error { return nil }) // Success
+		stats = cb.GetStats()
+
+		if stats.Failures != 0 {
+			t.Errorf("Expected 0 failures after success, got %d", stats.Failures)
+		}
+	})
+}
+
+func TestCircuitBreakerManager(t *testing.T) {
+	config := &Config{
+		LogLevel:                       logging.LogLevelError,
+		CircuitBreakerFailureThreshold: 5,
+		CircuitBreakerResetTimeout:     60 * time.Second,
+		CircuitBreakerMaxRequests:      3,
+	}
+
+	t.Run("GetCircuitBreaker", func(t *testing.T) {
+		manager := newCircuitBreakerManager(config)
+
+		cb1 := manager.GetCircuitBreaker("endpoint1:6379")
+		cb2 := manager.GetCircuitBreaker("endpoint2:6379")
+		cb3 := manager.GetCircuitBreaker("endpoint1:6379") // Same as cb1
+
+		if cb1 == cb2 {
+			t.Error("Different endpoints should have different circuit breakers")
+		}
+
+		if cb1 != cb3 {
+			t.Error("Same endpoint should return the same circuit breaker")
+		}
+	})
+
+	t.Run("GetAllStats", func(t *testing.T) {
+		manager := newCircuitBreakerManager(config)
+
+		// Create circuit breakers for different endpoints
+		cb1 := manager.GetCircuitBreaker("endpoint1:6379")
+		cb2 := manager.GetCircuitBreaker("endpoint2:6379")
+
+		// Execute some operations
+		cb1.Execute(func() error { return nil })
+		cb2.Execute(func() error { return errors.New("test error") })
+
+		stats := manager.GetAllStats()
+
+		if len(stats) != 2 {
+			t.Errorf("Expected 2 circuit breaker stats, got %d", len(stats))
+		}
+
+		// Check that we have stats for both endpoints
+		endpoints := make(map[string]bool)
+		for _, stat := range stats {
+			endpoints[stat.Endpoint] = true
+		}
+
+		if !endpoints["endpoint1:6379"] || !endpoints["endpoint2:6379"] {
+			t.Error("Missing stats for expected endpoints")
+		}
+	})
+
+	t.Run("Reset", func(t *testing.T) {
+		manager := newCircuitBreakerManager(config)
+		testError := errors.New("test error")
+
+		cb := manager.GetCircuitBreaker("test-endpoint:6379")
+
+		// Force circuit to open
+		for i := 0; i < 5; i++ {
+			cb.Execute(func() error { return testError })
+		}
+
+		if cb.GetState() != CircuitBreakerOpen {
+			t.Error("Circuit should be open")
+		}
+
+		// Reset all circuit breakers
+		manager.Reset()
+
+		if cb.GetState() != CircuitBreakerClosed {
+			t.Error("Circuit should be closed after reset")
+		}
+
+		if cb.failures.Load() != 0 {
+			t.Error("Failure count should be reset to 0")
+		}
+	})
+
+	t.Run("ConfigurableParameters", func(t *testing.T) {
+		config := &Config{
+			LogLevel:                       logging.LogLevelError,
+			CircuitBreakerFailureThreshold: 10,
+			CircuitBreakerResetTimeout:     30 * time.Second,
+			CircuitBreakerMaxRequests:      5,
+		}
+
+		cb := newCircuitBreaker("test-endpoint:6379", config)
+
+		// Test that configuration values are used
+		if cb.failureThreshold != 10 {
+			t.Errorf("Expected failureThreshold=10, got %d", cb.failureThreshold)
+		}
+		if cb.resetTimeout != 30*time.Second {
+			t.Errorf("Expected resetTimeout=30s, got %v", cb.resetTimeout)
+		}
+		if cb.maxRequests != 5 {
+			t.Errorf("Expected maxRequests=5, got %d", cb.maxRequests)
+		}
+
+		// Test that circuit opens after configured threshold
+		testError := errors.New("test error")
+		for i := 0; i < 9; i++ {
+			err := cb.Execute(func() error { return testError })
+			if err != testError {
+				t.Errorf("Expected test error, got %v", err)
+			}
+			if cb.GetState() != CircuitBreakerClosed {
+				t.Errorf("Circuit should still be closed after %d failures", i+1)
+			}
+		}
+
+		// 10th failure should open the circuit
+		err := cb.Execute(func() error { return testError })
+		if err != testError {
+			t.Errorf("Expected test error, got %v", err)
+		}
+
+		if cb.GetState() != CircuitBreakerOpen {
+			t.Errorf("Expected state %v, got %v", CircuitBreakerOpen, cb.GetState())
+		}
+	})
+}

hitless/config.go

@@ -0,0 +1,472 @@
+package hitless
+
+import (
+	"context"
+	"net"
+	"runtime"
+	"strings"
+	"time"
+
+	"github.com/redis/go-redis/v9/internal"
+	"github.com/redis/go-redis/v9/internal/util"
+	"github.com/redis/go-redis/v9/logging"
+)
+
+// MaintNotificationsMode represents the maintenance notifications mode
+type MaintNotificationsMode string
+
+// Constants for maintenance push notifications modes
+const (
+	MaintNotificationsDisabled MaintNotificationsMode = "disabled" // Client doesn't send CLIENT MAINT_NOTIFICATIONS ON command
+	MaintNotificationsEnabled  MaintNotificationsMode = "enabled"  // Client forcefully sends command, interrupts connection on error
+	MaintNotificationsAuto     MaintNotificationsMode = "auto"     // Client tries to send command, disables feature on error
+)
+
+// IsValid returns true if the maintenance notifications mode is valid
+func (m MaintNotificationsMode) IsValid() bool {
+	switch m {
+	case MaintNotificationsDisabled, MaintNotificationsEnabled, MaintNotificationsAuto:
+		return true
+	default:
+		return false
+	}
+}
+
+// String returns the string representation of the mode
+func (m MaintNotificationsMode) String() string {
+	return string(m)
+}
+
+// EndpointType represents the type of endpoint to request in MOVING notifications
+type EndpointType string
+
+// Constants for endpoint types
+const (
+	EndpointTypeAuto         EndpointType = "auto"          // Auto-detect based on connection
+	EndpointTypeInternalIP   EndpointType = "internal-ip"   // Internal IP address
+	EndpointTypeInternalFQDN EndpointType = "internal-fqdn" // Internal FQDN
+	EndpointTypeExternalIP   EndpointType = "external-ip"   // External IP address
+	EndpointTypeExternalFQDN EndpointType = "external-fqdn" // External FQDN
+	EndpointTypeNone         EndpointType = "none"          // No endpoint (reconnect with current config)
+)
+
+// IsValid returns true if the endpoint type is valid
+func (e EndpointType) IsValid() bool {
+	switch e {
+	case EndpointTypeAuto, EndpointTypeInternalIP, EndpointTypeInternalFQDN,
+		EndpointTypeExternalIP, EndpointTypeExternalFQDN, EndpointTypeNone:
+		return true
+	default:
+		return false
+	}
+}
+
+// String returns the string representation of the endpoint type
+func (e EndpointType) String() string {
+	return string(e)
+}
+
+// Config provides configuration options for hitless upgrades.
+type Config struct {
+	// Mode controls how client maintenance notifications are handled.
+	// Valid values: MaintNotificationsDisabled, MaintNotificationsEnabled, MaintNotificationsAuto
+	// Default: MaintNotificationsAuto
+	Mode MaintNotificationsMode
+
+	// EndpointType specifies the type of endpoint to request in MOVING notifications.
+	// Valid values: EndpointTypeAuto, EndpointTypeInternalIP, EndpointTypeInternalFQDN,
+	//               EndpointTypeExternalIP, EndpointTypeExternalFQDN, EndpointTypeNone
+	// Default: EndpointTypeAuto
+	EndpointType EndpointType
+
+	// RelaxedTimeout is the concrete timeout value to use during
+	// MIGRATING/FAILING_OVER states to accommodate increased latency.
+	// This applies to both read and write timeouts.
+	// Default: 10 seconds
+	RelaxedTimeout time.Duration
+
+	// HandoffTimeout is the maximum time to wait for connection handoff to complete.
+	// If handoff takes longer than this, the old connection will be forcibly closed.
+	// Default: 15 seconds (matches server-side eviction timeout)
+	HandoffTimeout time.Duration
+
+	// MaxWorkers is the maximum number of worker goroutines for processing handoff requests.
+	// Workers are created on-demand and automatically cleaned up when idle.
+	// If zero, defaults to min(10, PoolSize/2) to handle bursts effectively.
+	// If explicitly set, enforces minimum of PoolSize/2
+	//
+	// Default: min(PoolSize/2, max(10, PoolSize/3)), Minimum when set: PoolSize/2
+	MaxWorkers int
+
+	// HandoffQueueSize is the size of the buffered channel used to queue handoff requests.
+	// If the queue is full, new handoff requests will be rejected.
+	// Scales with both worker count and pool size for better burst handling.
+	//
+	// Default: max(20×MaxWorkers, PoolSize), capped by MaxActiveConns+1 (if set) or 5×PoolSize
+	// When set: minimum 200, capped by MaxActiveConns+1 (if set) or 5×PoolSize
+	HandoffQueueSize int
+
+	// PostHandoffRelaxedDuration is how long to keep relaxed timeouts on the new connection
+	// after a handoff completes. This provides additional resilience during cluster transitions.
+	// Default: 2 * RelaxedTimeout
+	PostHandoffRelaxedDuration time.Duration
+
+	// LogLevel controls the verbosity of hitless upgrade logging.
+	// LogLevelError (0) = errors only, LogLevelWarn (1) = warnings, LogLevelInfo (2) = info, LogLevelDebug (3) = debug
+	// Default: logging.LogLevelError(0)
+	LogLevel logging.LogLevel
+
+	// Circuit breaker configuration for endpoint failure handling
+	// CircuitBreakerFailureThreshold is the number of failures before opening the circuit.
+	// Default: 5
+	CircuitBreakerFailureThreshold int
+
+	// CircuitBreakerResetTimeout is how long to wait before testing if the endpoint recovered.
+	// Default: 60 seconds
+	CircuitBreakerResetTimeout time.Duration
+
+	// CircuitBreakerMaxRequests is the maximum number of requests allowed in half-open state.
+	// Default: 3
+	CircuitBreakerMaxRequests int
+
+	// MaxHandoffRetries is the maximum number of times to retry a failed handoff.
+	// After this many retries, the connection will be removed from the pool.
+	// Default: 3
+	MaxHandoffRetries int
+}
+
+func (c *Config) IsEnabled() bool {
+	return c != nil && c.Mode != MaintNotificationsDisabled
+}
+
+// DefaultConfig returns a Config with sensible defaults.
+func DefaultConfig() *Config {
+	return &Config{
+		Mode:                       MaintNotificationsAuto, // Enable by default for Redis Cloud
+		EndpointType:               EndpointTypeAuto,       // Auto-detect based on connection
+		RelaxedTimeout:             10 * time.Second,
+		HandoffTimeout:             15 * time.Second,
+		MaxWorkers:                 0, // Auto-calculated based on pool size
+		HandoffQueueSize:           0, // Auto-calculated based on max workers
+		PostHandoffRelaxedDuration: 0, // Auto-calculated based on relaxed timeout
+		LogLevel:                   logging.LogLevelError,
+
+		// Circuit breaker configuration
+		CircuitBreakerFailureThreshold: 5,
+		CircuitBreakerResetTimeout:     60 * time.Second,
+		CircuitBreakerMaxRequests:      3,
+
+		// Connection Handoff Configuration
+		MaxHandoffRetries: 3,
+	}
+}
+
+// Validate checks if the configuration is valid.
+func (c *Config) Validate() error {
+	if c.RelaxedTimeout <= 0 {
+		return ErrInvalidRelaxedTimeout
+	}
+	if c.HandoffTimeout <= 0 {
+		return ErrInvalidHandoffTimeout
+	}
+	// Validate worker configuration
+	// Allow 0 for auto-calculation, but negative values are invalid
+	if c.MaxWorkers < 0 {
+		return ErrInvalidHandoffWorkers
+	}
+	// HandoffQueueSize validation - allow 0 for auto-calculation
+	if c.HandoffQueueSize < 0 {
+		return ErrInvalidHandoffQueueSize
+	}
+	if c.PostHandoffRelaxedDuration < 0 {
+		return ErrInvalidPostHandoffRelaxedDuration
+	}
+	if !c.LogLevel.IsValid() {
+		return ErrInvalidLogLevel
+	}
+
+	// Circuit breaker validation
+	if c.CircuitBreakerFailureThreshold < 1 {
+		return ErrInvalidCircuitBreakerFailureThreshold
+	}
+	if c.CircuitBreakerResetTimeout < 0 {
+		return ErrInvalidCircuitBreakerResetTimeout
+	}
+	if c.CircuitBreakerMaxRequests < 1 {
+		return ErrInvalidCircuitBreakerMaxRequests
+	}
+
+	// Validate Mode (maintenance notifications mode)
+	if !c.Mode.IsValid() {
+		return ErrInvalidMaintNotifications
+	}
+
+	// Validate EndpointType
+	if !c.EndpointType.IsValid() {
+		return ErrInvalidEndpointType
+	}
+
+	// Validate configuration fields
+	if c.MaxHandoffRetries < 1 || c.MaxHandoffRetries > 10 {
+		return ErrInvalidHandoffRetries
+	}
+
+	return nil
+}
+
+// ApplyDefaults applies default values to any zero-value fields in the configuration.
+// This ensures that partially configured structs get sensible defaults for missing fields.
+func (c *Config) ApplyDefaults() *Config {
+	return c.ApplyDefaultsWithPoolSize(0)
+}
+
+// ApplyDefaultsWithPoolSize applies default values to any zero-value fields in the configuration,
+// using the provided pool size to calculate worker defaults.
+// This ensures that partially configured structs get sensible defaults for missing fields.
+func (c *Config) ApplyDefaultsWithPoolSize(poolSize int) *Config {
+	return c.ApplyDefaultsWithPoolConfig(poolSize, 0)
+}
+
+// ApplyDefaultsWithPoolConfig applies default values to any zero-value fields in the configuration,
+// using the provided pool size and max active connections to calculate worker and queue defaults.
+// This ensures that partially configured structs get sensible defaults for missing fields.
+func (c *Config) ApplyDefaultsWithPoolConfig(poolSize int, maxActiveConns int) *Config {
+	if c == nil {
+		return DefaultConfig().ApplyDefaultsWithPoolSize(poolSize)
+	}
+
+	defaults := DefaultConfig()
+	result := &Config{}
+
+	// Apply defaults for enum fields (empty/zero means not set)
+	result.Mode = defaults.Mode
+	if c.Mode != "" {
+		result.Mode = c.Mode
+	}
+
+	result.EndpointType = defaults.EndpointType
+	if c.EndpointType != "" {
+		result.EndpointType = c.EndpointType
+	}
+
+	// Apply defaults for duration fields (zero means not set)
+	result.RelaxedTimeout = defaults.RelaxedTimeout
+	if c.RelaxedTimeout > 0 {
+		result.RelaxedTimeout = c.RelaxedTimeout
+	}
+
+	result.HandoffTimeout = defaults.HandoffTimeout
+	if c.HandoffTimeout > 0 {
+		result.HandoffTimeout = c.HandoffTimeout
+	}
+
+	// Copy worker configuration
+	result.MaxWorkers = c.MaxWorkers
+
+	// Apply worker defaults based on pool size
+	result.applyWorkerDefaults(poolSize)
+
+	// Apply queue size defaults with new scaling approach
+	// Default: max(20x workers, PoolSize), capped by maxActiveConns or 5x pool size
+	workerBasedSize := result.MaxWorkers * 20
+	poolBasedSize := poolSize
+	result.HandoffQueueSize = util.Max(workerBasedSize, poolBasedSize)
+	if c.HandoffQueueSize > 0 {
+		// When explicitly set: enforce minimum of 200
+		result.HandoffQueueSize = util.Max(200, c.HandoffQueueSize)
+	}
+
+	// Cap queue size: use maxActiveConns+1 if set, otherwise 5x pool size
+	var queueCap int
+	if maxActiveConns > 0 {
+		queueCap = maxActiveConns + 1
+		// Ensure queue cap is at least 2 for very small maxActiveConns
+		if queueCap < 2 {
+			queueCap = 2
+		}
+	} else {
+		queueCap = poolSize * 5
+	}
+	result.HandoffQueueSize = util.Min(result.HandoffQueueSize, queueCap)
+
+	// Ensure minimum queue size of 2 (fallback for very small pools)
+	if result.HandoffQueueSize < 2 {
+		result.HandoffQueueSize = 2
+	}
+
+	result.PostHandoffRelaxedDuration = result.RelaxedTimeout * 2
+	if c.PostHandoffRelaxedDuration > 0 {
+		result.PostHandoffRelaxedDuration = c.PostHandoffRelaxedDuration
+	}
+
+	// LogLevel: 0 is a valid value (errors only), so we need to check if it was explicitly set
+	// We'll use the provided value as-is, since 0 is valid
+	result.LogLevel = c.LogLevel
+
+	// Apply defaults for configuration fields
+	result.MaxHandoffRetries = defaults.MaxHandoffRetries
+	if c.MaxHandoffRetries > 0 {
+		result.MaxHandoffRetries = c.MaxHandoffRetries
+	}
+
+	// Circuit breaker configuration
+	result.CircuitBreakerFailureThreshold = defaults.CircuitBreakerFailureThreshold
+	if c.CircuitBreakerFailureThreshold > 0 {
+		result.CircuitBreakerFailureThreshold = c.CircuitBreakerFailureThreshold
+	}
+
+	result.CircuitBreakerResetTimeout = defaults.CircuitBreakerResetTimeout
+	if c.CircuitBreakerResetTimeout > 0 {
+		result.CircuitBreakerResetTimeout = c.CircuitBreakerResetTimeout
+	}
+
+	result.CircuitBreakerMaxRequests = defaults.CircuitBreakerMaxRequests
+	if c.CircuitBreakerMaxRequests > 0 {
+		result.CircuitBreakerMaxRequests = c.CircuitBreakerMaxRequests
+	}
+
+	if result.LogLevel.DebugOrAbove() {
+		internal.Logger.Printf(context.Background(), "hitless: debug logging enabled")
+		internal.Logger.Printf(context.Background(), "hitless: config: %+v", result)
+	}
+	return result
+}
+
+// Clone creates a deep copy of the configuration.
+func (c *Config) Clone() *Config {
+	if c == nil {
+		return DefaultConfig()
+	}
+
+	return &Config{
+		Mode:                       c.Mode,
+		EndpointType:               c.EndpointType,
+		RelaxedTimeout:             c.RelaxedTimeout,
+		HandoffTimeout:             c.HandoffTimeout,
+		MaxWorkers:                 c.MaxWorkers,
+		HandoffQueueSize:           c.HandoffQueueSize,
+		PostHandoffRelaxedDuration: c.PostHandoffRelaxedDuration,
+		LogLevel:                   c.LogLevel,
+
+		// Circuit breaker configuration
+		CircuitBreakerFailureThreshold: c.CircuitBreakerFailureThreshold,
+		CircuitBreakerResetTimeout:     c.CircuitBreakerResetTimeout,
+		CircuitBreakerMaxRequests:      c.CircuitBreakerMaxRequests,
+
+		// Configuration fields
+		MaxHandoffRetries: c.MaxHandoffRetries,
+	}
+}
+
+// applyWorkerDefaults calculates and applies worker defaults based on pool size
+func (c *Config) applyWorkerDefaults(poolSize int) {
+	// Calculate defaults based on pool size
+	if poolSize <= 0 {
+		poolSize = 10 * runtime.GOMAXPROCS(0)
+	}
+
+	// When not set: min(poolSize/2, max(10, poolSize/3)) - balanced scaling approach
+	originalMaxWorkers := c.MaxWorkers
+	c.MaxWorkers = util.Min(poolSize/2, util.Max(10, poolSize/3))
+	if originalMaxWorkers != 0 {
+		// When explicitly set: max(poolSize/2, set_value) - ensure at least poolSize/2 workers
+		c.MaxWorkers = util.Max(poolSize/2, originalMaxWorkers)
+	}
+
+	// Ensure minimum of 1 worker (fallback for very small pools)
+	if c.MaxWorkers < 1 {
+		c.MaxWorkers = 1
+	}
+}
+
+// DetectEndpointType automatically detects the appropriate endpoint type
+// based on the connection address and TLS configuration.
+//
+// For IP addresses:
+//   - If TLS is enabled: requests FQDN for proper certificate validation
+//   - If TLS is disabled: requests IP for better performance
+//
+// For hostnames:
+//   - If TLS is enabled: always requests FQDN for proper certificate validation
+//   - If TLS is disabled: requests IP for better performance
+//
+// Internal vs External detection:
+//   - For IPs: uses private IP range detection
+//   - For hostnames: uses heuristics based on common internal naming patterns
+func DetectEndpointType(addr string, tlsEnabled bool) EndpointType {
+	// Extract host from "host:port" format
+	host, _, err := net.SplitHostPort(addr)
+	if err != nil {
+		host = addr // Assume no port
+	}
+
+	// Check if the host is an IP address or hostname
+	ip := net.ParseIP(host)
+	isIPAddress := ip != nil
+	var endpointType EndpointType
+
+	if isIPAddress {
+		// Address is an IP - determine if it's private or public
+		isPrivate := ip.IsPrivate() || ip.IsLoopback() || ip.IsLinkLocalUnicast()
+
+		if tlsEnabled {
+			// TLS with IP addresses - still prefer FQDN for certificate validation
+			if isPrivate {
+				endpointType = EndpointTypeInternalFQDN
+			} else {
+				endpointType = EndpointTypeExternalFQDN
+			}
+		} else {
+			// No TLS - can use IP addresses directly
+			if isPrivate {
+				endpointType = EndpointTypeInternalIP
+			} else {
+				endpointType = EndpointTypeExternalIP
+			}
+		}
+	} else {
+		// Address is a hostname
+		isInternalHostname := isInternalHostname(host)
+		if isInternalHostname {
+			endpointType = EndpointTypeInternalFQDN
+		} else {
+			endpointType = EndpointTypeExternalFQDN
+		}
+	}
+
+	return endpointType
+}
+
+// isInternalHostname determines if a hostname appears to be internal/private.
+// This is a heuristic based on common naming patterns.
+func isInternalHostname(hostname string) bool {
+	// Convert to lowercase for comparison
+	hostname = strings.ToLower(hostname)
+
+	// Common internal hostname patterns
+	internalPatterns := []string{
+		"localhost",
+		".local",
+		".internal",
+		".corp",
+		".lan",
+		".intranet",
+		".private",
+	}
+
+	// Check for exact match or suffix match
+	for _, pattern := range internalPatterns {
+		if hostname == pattern || strings.HasSuffix(hostname, pattern) {
+			return true
+		}
+	}
+
+	// Check for RFC 1918 style hostnames (e.g., redis-1, db-server, etc.)
+	// If hostname doesn't contain dots, it's likely internal
+	if !strings.Contains(hostname, ".") {
+		return true
+	}
+
+	// Default to external for fully qualified domain names
+	return false
+}

hitless/config_test.go

@@ -0,0 +1,490 @@
+package hitless
+
+import (
+	"context"
+	"net"
+	"testing"
+	"time"
+
+	"github.com/redis/go-redis/v9/internal/util"
+	"github.com/redis/go-redis/v9/logging"
+)
+
+func TestConfig(t *testing.T) {
+	t.Run("DefaultConfig", func(t *testing.T) {
+		config := DefaultConfig()
+
+		// MaxWorkers should be 0 in default config (auto-calculated)
+		if config.MaxWorkers != 0 {
+			t.Errorf("Expected MaxWorkers to be 0 (auto-calculated), got %d", config.MaxWorkers)
+		}
+
+		// HandoffQueueSize should be 0 in default config (auto-calculated)
+		if config.HandoffQueueSize != 0 {
+			t.Errorf("Expected HandoffQueueSize to be 0 (auto-calculated), got %d", config.HandoffQueueSize)
+		}
+
+		if config.RelaxedTimeout != 10*time.Second {
+			t.Errorf("Expected RelaxedTimeout to be 10s, got %v", config.RelaxedTimeout)
+		}
+
+		// Test configuration fields have proper defaults
+		if config.MaxHandoffRetries != 3 {
+			t.Errorf("Expected MaxHandoffRetries to be 3, got %d", config.MaxHandoffRetries)
+		}
+
+		// Circuit breaker defaults
+		if config.CircuitBreakerFailureThreshold != 5 {
+			t.Errorf("Expected CircuitBreakerFailureThreshold=5, got %d", config.CircuitBreakerFailureThreshold)
+		}
+		if config.CircuitBreakerResetTimeout != 60*time.Second {
+			t.Errorf("Expected CircuitBreakerResetTimeout=60s, got %v", config.CircuitBreakerResetTimeout)
+		}
+		if config.CircuitBreakerMaxRequests != 3 {
+			t.Errorf("Expected CircuitBreakerMaxRequests=3, got %d", config.CircuitBreakerMaxRequests)
+		}
+
+		if config.HandoffTimeout != 15*time.Second {
+			t.Errorf("Expected HandoffTimeout to be 15s, got %v", config.HandoffTimeout)
+		}
+
+		if config.PostHandoffRelaxedDuration != 0 {
+			t.Errorf("Expected PostHandoffRelaxedDuration to be 0 (auto-calculated), got %v", config.PostHandoffRelaxedDuration)
+		}
+
+		// Test that defaults are applied correctly
+		configWithDefaults := config.ApplyDefaultsWithPoolSize(100)
+		if configWithDefaults.PostHandoffRelaxedDuration != 20*time.Second {
+			t.Errorf("Expected PostHandoffRelaxedDuration to be 20s (2x RelaxedTimeout) after applying defaults, got %v", configWithDefaults.PostHandoffRelaxedDuration)
+		}
+	})
+
+	t.Run("ConfigValidation", func(t *testing.T) {
+		// Valid config with applied defaults
+		config := DefaultConfig().ApplyDefaults()
+		if err := config.Validate(); err != nil {
+			t.Errorf("Default config with applied defaults should be valid: %v", err)
+		}
+
+		// Invalid worker configuration (negative MaxWorkers)
+		config = &Config{
+			RelaxedTimeout:             30 * time.Second,
+			HandoffTimeout:             15 * time.Second,
+			MaxWorkers:                 -1, // This should be invalid
+			HandoffQueueSize:           100,
+			PostHandoffRelaxedDuration: 10 * time.Second,
+			LogLevel:                   1,
+			MaxHandoffRetries:          3, // Add required field
+		}
+		if err := config.Validate(); err != ErrInvalidHandoffWorkers {
+			t.Errorf("Expected ErrInvalidHandoffWorkers, got %v", err)
+		}
+
+		// Invalid HandoffQueueSize
+		config = DefaultConfig().ApplyDefaults()
+		config.HandoffQueueSize = -1
+		if err := config.Validate(); err != ErrInvalidHandoffQueueSize {
+			t.Errorf("Expected ErrInvalidHandoffQueueSize, got %v", err)
+		}
+
+		// Invalid PostHandoffRelaxedDuration
+		config = DefaultConfig().ApplyDefaults()
+		config.PostHandoffRelaxedDuration = -1 * time.Second
+		if err := config.Validate(); err != ErrInvalidPostHandoffRelaxedDuration {
+			t.Errorf("Expected ErrInvalidPostHandoffRelaxedDuration, got %v", err)
+		}
+	})
+
+	t.Run("ConfigClone", func(t *testing.T) {
+		original := DefaultConfig()
+		original.MaxWorkers = 20
+		original.HandoffQueueSize = 200
+
+		cloned := original.Clone()
+
+		if cloned.MaxWorkers != 20 {
+			t.Errorf("Expected cloned MaxWorkers to be 20, got %d", cloned.MaxWorkers)
+		}
+
+		if cloned.HandoffQueueSize != 200 {
+			t.Errorf("Expected cloned HandoffQueueSize to be 200, got %d", cloned.HandoffQueueSize)
+		}
+
+		// Modify original to ensure clone is independent
+		original.MaxWorkers = 2
+		if cloned.MaxWorkers != 20 {
+			t.Error("Clone should be independent of original")
+		}
+	})
+}
+
+func TestApplyDefaults(t *testing.T) {
+	t.Run("NilConfig", func(t *testing.T) {
+		var config *Config
+		result := config.ApplyDefaultsWithPoolSize(100) // Use explicit pool size for testing
+
+		// With nil config, should get default config with auto-calculated workers
+		if result.MaxWorkers <= 0 {
+			t.Errorf("Expected MaxWorkers to be > 0 after applying defaults, got %d", result.MaxWorkers)
+		}
+
+		// HandoffQueueSize should be auto-calculated with hybrid scaling
+		workerBasedSize := result.MaxWorkers * 20
+		poolSize := 100 // Default pool size used in ApplyDefaults
+		poolBasedSize := poolSize
+		expectedQueueSize := util.Max(workerBasedSize, poolBasedSize)
+		expectedQueueSize = util.Min(expectedQueueSize, poolSize*5) // Cap by 5x pool size
+		if result.HandoffQueueSize != expectedQueueSize {
+			t.Errorf("Expected HandoffQueueSize to be %d (max(20*MaxWorkers=%d, poolSize=%d) capped by 5*poolSize=%d), got %d",
+				expectedQueueSize, workerBasedSize, poolBasedSize, poolSize*5, result.HandoffQueueSize)
+		}
+	})
+
+	t.Run("PartialConfig", func(t *testing.T) {
+		config := &Config{
+			MaxWorkers: 60, // Set this field explicitly (> poolSize/2 = 50)
+			// Leave other fields as zero values
+		}
+
+		result := config.ApplyDefaultsWithPoolSize(100) // Use explicit pool size for testing
+
+		// Should keep the explicitly set values when > poolSize/2
+		if result.MaxWorkers != 60 {
+			t.Errorf("Expected MaxWorkers to be 60 (explicitly set), got %d", result.MaxWorkers)
+		}
+
+		// Should apply default for unset fields (auto-calculated queue size with hybrid scaling)
+		workerBasedSize := result.MaxWorkers * 20
+		poolSize := 100 // Default pool size used in ApplyDefaults
+		poolBasedSize := poolSize
+		expectedQueueSize := util.Max(workerBasedSize, poolBasedSize)
+		expectedQueueSize = util.Min(expectedQueueSize, poolSize*5) // Cap by 5x pool size
+		if result.HandoffQueueSize != expectedQueueSize {
+			t.Errorf("Expected HandoffQueueSize to be %d (max(20*MaxWorkers=%d, poolSize=%d) capped by 5*poolSize=%d), got %d",
+				expectedQueueSize, workerBasedSize, poolBasedSize, poolSize*5, result.HandoffQueueSize)
+		}
+
+		// Test explicit queue size capping by 5x pool size
+		configWithLargeQueue := &Config{
+			MaxWorkers:       5,
+			HandoffQueueSize: 1000, // Much larger than 5x pool size
+		}
+
+		resultCapped := configWithLargeQueue.ApplyDefaultsWithPoolSize(20) // Small pool size
+		expectedCap := 20 * 5                                              // 5x pool size = 100
+		if resultCapped.HandoffQueueSize != expectedCap {
+			t.Errorf("Expected HandoffQueueSize to be capped by 5x pool size (%d), got %d", expectedCap, resultCapped.HandoffQueueSize)
+		}
+
+		// Test explicit queue size minimum enforcement
+		configWithSmallQueue := &Config{
+			MaxWorkers:       5,
+			HandoffQueueSize: 10, // Below minimum of 200
+		}
+
+		resultMinimum := configWithSmallQueue.ApplyDefaultsWithPoolSize(100) // Large pool size
+		if resultMinimum.HandoffQueueSize != 200 {
+			t.Errorf("Expected HandoffQueueSize to be enforced minimum (200), got %d", resultMinimum.HandoffQueueSize)
+		}
+
+		// Test that large explicit values are capped by 5x pool size
+		configWithVeryLargeQueue := &Config{
+			MaxWorkers:       5,
+			HandoffQueueSize: 1000, // Much larger than 5x pool size
+		}
+
+		resultVeryLarge := configWithVeryLargeQueue.ApplyDefaultsWithPoolSize(100) // Pool size 100
+		expectedVeryLargeCap := 100 * 5                                            // 5x pool size = 500
+		if resultVeryLarge.HandoffQueueSize != expectedVeryLargeCap {
+			t.Errorf("Expected very large HandoffQueueSize to be capped by 5x pool size (%d), got %d", expectedVeryLargeCap, resultVeryLarge.HandoffQueueSize)
+		}
+
+		if result.RelaxedTimeout != 10*time.Second {
+			t.Errorf("Expected RelaxedTimeout to be 10s (default), got %v", result.RelaxedTimeout)
+		}
+
+		if result.HandoffTimeout != 15*time.Second {
+			t.Errorf("Expected HandoffTimeout to be 15s (default), got %v", result.HandoffTimeout)
+		}
+	})
+
+	t.Run("ZeroValues", func(t *testing.T) {
+		config := &Config{
+			MaxWorkers:       0, // Zero value should get auto-calculated defaults
+			HandoffQueueSize: 0, // Zero value should get default
+			RelaxedTimeout:   0, // Zero value should get default
+			LogLevel:         0, // Zero is valid for LogLevel (errors only)
+		}
+
+		result := config.ApplyDefaultsWithPoolSize(100) // Use explicit pool size for testing
+
+		// Zero values should get auto-calculated defaults
+		if result.MaxWorkers <= 0 {
+			t.Errorf("Expected MaxWorkers to be > 0 (auto-calculated), got %d", result.MaxWorkers)
+		}
+
+		// HandoffQueueSize should be auto-calculated with hybrid scaling
+		workerBasedSize := result.MaxWorkers * 20
+		poolSize := 100 // Default pool size used in ApplyDefaults
+		poolBasedSize := poolSize
+		expectedQueueSize := util.Max(workerBasedSize, poolBasedSize)
+		expectedQueueSize = util.Min(expectedQueueSize, poolSize*5) // Cap by 5x pool size
+		if result.HandoffQueueSize != expectedQueueSize {
+			t.Errorf("Expected HandoffQueueSize to be %d (max(20*MaxWorkers=%d, poolSize=%d) capped by 5*poolSize=%d), got %d",
+				expectedQueueSize, workerBasedSize, poolBasedSize, poolSize*5, result.HandoffQueueSize)
+		}
+
+		if result.RelaxedTimeout != 10*time.Second {
+			t.Errorf("Expected RelaxedTimeout to be 10s (default), got %v", result.RelaxedTimeout)
+		}
+
+		// LogLevel 0 should be preserved (it's a valid value)
+		if result.LogLevel != 0 {
+			t.Errorf("Expected LogLevel to be 0 (preserved), got %d", result.LogLevel)
+		}
+	})
+}
+
+func TestProcessorWithConfig(t *testing.T) {
+	t.Run("ProcessorUsesConfigValues", func(t *testing.T) {
+		config := &Config{
+			MaxWorkers:       5,
+			HandoffQueueSize: 50,
+			RelaxedTimeout:   10 * time.Second,
+			HandoffTimeout:   5 * time.Second,
+		}
+
+		baseDialer := func(ctx context.Context, network, addr string) (net.Conn, error) {
+			return &mockNetConn{addr: addr}, nil
+		}
+
+		processor := NewPoolHook(baseDialer, "tcp", config, nil)
+		defer processor.Shutdown(context.Background())
+
+		// The processor should be created successfully with custom config
+		if processor == nil {
+			t.Error("Processor should be created with custom config")
+		}
+	})
+
+	t.Run("ProcessorWithPartialConfig", func(t *testing.T) {
+		config := &Config{
+			MaxWorkers: 7, // Only set worker field
+			// Other fields will get defaults
+		}
+
+		baseDialer := func(ctx context.Context, network, addr string) (net.Conn, error) {
+			return &mockNetConn{addr: addr}, nil
+		}
+
+		processor := NewPoolHook(baseDialer, "tcp", config, nil)
+		defer processor.Shutdown(context.Background())
+
+		// Should work with partial config (defaults applied)
+		if processor == nil {
+			t.Error("Processor should be created with partial config")
+		}
+	})
+
+	t.Run("ProcessorWithNilConfig", func(t *testing.T) {
+		baseDialer := func(ctx context.Context, network, addr string) (net.Conn, error) {
+			return &mockNetConn{addr: addr}, nil
+		}
+
+		processor := NewPoolHook(baseDialer, "tcp", nil, nil)
+		defer processor.Shutdown(context.Background())
+
+		// Should use default config when nil is passed
+		if processor == nil {
+			t.Error("Processor should be created with nil config (using defaults)")
+		}
+	})
+}
+
+func TestIntegrationWithApplyDefaults(t *testing.T) {
+	t.Run("ProcessorWithPartialConfigAppliesDefaults", func(t *testing.T) {
+		// Create a partial config with only some fields set
+		partialConfig := &Config{
+			MaxWorkers: 15,                   // Custom value (>= 10 to test preservation)
+			LogLevel:   logging.LogLevelInfo, // Custom value
+			// Other fields left as zero values - should get defaults
+		}
+
+		baseDialer := func(ctx context.Context, network, addr string) (net.Conn, error) {
+			return &mockNetConn{addr: addr}, nil
+		}
+
+		// Create processor - should apply defaults to missing fields
+		processor := NewPoolHook(baseDialer, "tcp", partialConfig, nil)
+		defer processor.Shutdown(context.Background())
+
+		// Processor should be created successfully
+		if processor == nil {
+			t.Error("Processor should be created with partial config")
+		}
+
+		// Test that the ApplyDefaults method worked correctly by creating the same config
+		// and applying defaults manually
+		expectedConfig := partialConfig.ApplyDefaultsWithPoolSize(100) // Use explicit pool size for testing
+
+		// Should preserve custom values (when >= poolSize/2)
+		if expectedConfig.MaxWorkers != 50 { // max(poolSize/2, 15) = max(50, 15) = 50
+			t.Errorf("Expected MaxWorkers to be 50, got %d", expectedConfig.MaxWorkers)
+		}
+
+		if expectedConfig.LogLevel != 2 {
+			t.Errorf("Expected LogLevel to be 2, got %d", expectedConfig.LogLevel)
+		}
+
+		// Should apply defaults for missing fields (auto-calculated queue size with hybrid scaling)
+		workerBasedSize := expectedConfig.MaxWorkers * 20
+		poolSize := 100 // Default pool size used in ApplyDefaults
+		poolBasedSize := poolSize
+		expectedQueueSize := util.Max(workerBasedSize, poolBasedSize)
+		expectedQueueSize = util.Min(expectedQueueSize, poolSize*5) // Cap by 5x pool size
+		if expectedConfig.HandoffQueueSize != expectedQueueSize {
+			t.Errorf("Expected HandoffQueueSize to be %d (max(20*MaxWorkers=%d, poolSize=%d) capped by 5*poolSize=%d), got %d",
+				expectedQueueSize, workerBasedSize, poolBasedSize, poolSize*5, expectedConfig.HandoffQueueSize)
+		}
+
+		// Test that queue size is always capped by 5x pool size
+		if expectedConfig.HandoffQueueSize > poolSize*5 {
+			t.Errorf("HandoffQueueSize (%d) should never exceed 5x pool size (%d)",
+				expectedConfig.HandoffQueueSize, poolSize*2)
+		}
+
+		if expectedConfig.RelaxedTimeout != 10*time.Second {
+			t.Errorf("Expected RelaxedTimeout to be 10s (default), got %v", expectedConfig.RelaxedTimeout)
+		}
+
+		if expectedConfig.HandoffTimeout != 15*time.Second {
+			t.Errorf("Expected HandoffTimeout to be 15s (default), got %v", expectedConfig.HandoffTimeout)
+		}
+
+		if expectedConfig.PostHandoffRelaxedDuration != 20*time.Second {
+			t.Errorf("Expected PostHandoffRelaxedDuration to be 20s (2x RelaxedTimeout), got %v", expectedConfig.PostHandoffRelaxedDuration)
+		}
+	})
+}
+
+func TestEnhancedConfigValidation(t *testing.T) {
+	t.Run("ValidateFields", func(t *testing.T) {
+		config := DefaultConfig()
+		config.ApplyDefaultsWithPoolSize(100) // Apply defaults with pool size 100
+
+		// Should pass validation with default values
+		if err := config.Validate(); err != nil {
+			t.Errorf("Default config should be valid, got error: %v", err)
+		}
+
+		// Test invalid MaxHandoffRetries
+		config.MaxHandoffRetries = 0
+		if err := config.Validate(); err == nil {
+			t.Error("Expected validation error for MaxHandoffRetries = 0")
+		}
+		config.MaxHandoffRetries = 11
+		if err := config.Validate(); err == nil {
+			t.Error("Expected validation error for MaxHandoffRetries = 11")
+		}
+		config.MaxHandoffRetries = 3 // Reset to valid value
+
+		// Test circuit breaker validation
+		config.CircuitBreakerFailureThreshold = 0
+		if err := config.Validate(); err != ErrInvalidCircuitBreakerFailureThreshold {
+			t.Errorf("Expected ErrInvalidCircuitBreakerFailureThreshold, got %v", err)
+		}
+		config.CircuitBreakerFailureThreshold = 5 // Reset to valid value
+
+		config.CircuitBreakerResetTimeout = -1 * time.Second
+		if err := config.Validate(); err != ErrInvalidCircuitBreakerResetTimeout {
+			t.Errorf("Expected ErrInvalidCircuitBreakerResetTimeout, got %v", err)
+		}
+		config.CircuitBreakerResetTimeout = 60 * time.Second // Reset to valid value
+
+		config.CircuitBreakerMaxRequests = 0
+		if err := config.Validate(); err != ErrInvalidCircuitBreakerMaxRequests {
+			t.Errorf("Expected ErrInvalidCircuitBreakerMaxRequests, got %v", err)
+		}
+		config.CircuitBreakerMaxRequests = 3 // Reset to valid value
+
+		// Should pass validation again
+		if err := config.Validate(); err != nil {
+			t.Errorf("Config should be valid after reset, got error: %v", err)
+		}
+	})
+}
+
+func TestConfigClone(t *testing.T) {
+	original := DefaultConfig()
+	original.MaxHandoffRetries = 7
+	original.HandoffTimeout = 8 * time.Second
+
+	cloned := original.Clone()
+
+	// Test that values are copied
+	if cloned.MaxHandoffRetries != 7 {
+		t.Errorf("Expected cloned MaxHandoffRetries to be 7, got %d", cloned.MaxHandoffRetries)
+	}
+	if cloned.HandoffTimeout != 8*time.Second {
+		t.Errorf("Expected cloned HandoffTimeout to be 8s, got %v", cloned.HandoffTimeout)
+	}
+
+	// Test that modifying clone doesn't affect original
+	cloned.MaxHandoffRetries = 10
+	if original.MaxHandoffRetries != 7 {
+		t.Errorf("Modifying clone should not affect original, original MaxHandoffRetries changed to %d", original.MaxHandoffRetries)
+	}
+}
+
+func TestMaxWorkersLogic(t *testing.T) {
+	t.Run("AutoCalculatedMaxWorkers", func(t *testing.T) {
+		testCases := []struct {
+			poolSize        int
+			expectedWorkers int
+			description     string
+		}{
+			{6, 3, "Small pool: min(6/2, max(10, 6/3)) = min(3, max(10, 2)) = min(3, 10) = 3"},
+			{15, 7, "Medium pool: min(15/2, max(10, 15/3)) = min(7, max(10, 5)) = min(7, 10) = 7"},
+			{30, 10, "Large pool: min(30/2, max(10, 30/3)) = min(15, max(10, 10)) = min(15, 10) = 10"},
+			{60, 20, "Very large pool: min(60/2, max(10, 60/3)) = min(30, max(10, 20)) = min(30, 20) = 20"},
+			{120, 40, "Huge pool: min(120/2, max(10, 120/3)) = min(60, max(10, 40)) = min(60, 40) = 40"},
+		}
+
+		for _, tc := range testCases {
+			config := &Config{} // MaxWorkers = 0 (not set)
+			result := config.ApplyDefaultsWithPoolSize(tc.poolSize)
+
+			if result.MaxWorkers != tc.expectedWorkers {
+				t.Errorf("PoolSize=%d: expected MaxWorkers=%d, got %d (%s)",
+					tc.poolSize, tc.expectedWorkers, result.MaxWorkers, tc.description)
+			}
+		}
+	})
+
+	t.Run("ExplicitlySetMaxWorkers", func(t *testing.T) {
+		testCases := []struct {
+			setValue        int
+			expectedWorkers int
+			description     string
+		}{
+			{1, 50, "Set 1: max(poolSize/2, 1) = max(50, 1) = 50 (enforced minimum)"},
+			{5, 50, "Set 5: max(poolSize/2, 5) = max(50, 5) = 50 (enforced minimum)"},
+			{8, 50, "Set 8: max(poolSize/2, 8) = max(50, 8) = 50 (enforced minimum)"},
+			{10, 50, "Set 10: max(poolSize/2, 10) = max(50, 10) = 50 (enforced minimum)"},
+			{15, 50, "Set 15: max(poolSize/2, 15) = max(50, 15) = 50 (enforced minimum)"},
+			{60, 60, "Set 60: max(poolSize/2, 60) = max(50, 60) = 60 (respects user choice)"},
+		}
+
+		for _, tc := range testCases {
+			config := &Config{
+				MaxWorkers: tc.setValue, // Explicitly set
+			}
+			result := config.ApplyDefaultsWithPoolSize(100) // Pool size doesn't affect explicit values
+
+			if result.MaxWorkers != tc.expectedWorkers {
+				t.Errorf("Set MaxWorkers=%d: expected %d, got %d (%s)",
+					tc.setValue, tc.expectedWorkers, result.MaxWorkers, tc.description)
+			}
+		}
+	})
+}

hitless/errors.go

@@ -0,0 +1,105 @@
+package hitless
+
+import (
+	"errors"
+	"fmt"
+	"time"
+)
+
+// Configuration errors
+var (
+	ErrInvalidRelaxedTimeout             = errors.New("hitless: relaxed timeout must be greater than 0")
+	ErrInvalidHandoffTimeout             = errors.New("hitless: handoff timeout must be greater than 0")
+	ErrInvalidHandoffWorkers             = errors.New("hitless: MaxWorkers must be greater than or equal to 0")
+	ErrInvalidHandoffQueueSize           = errors.New("hitless: handoff queue size must be greater than 0")
+	ErrInvalidPostHandoffRelaxedDuration = errors.New("hitless: post-handoff relaxed duration must be greater than or equal to 0")
+	ErrInvalidLogLevel                   = errors.New("hitless: log level must be LogLevelError (0), LogLevelWarn (1), LogLevelInfo (2), or LogLevelDebug (3)")
+	ErrInvalidEndpointType               = errors.New("hitless: invalid endpoint type")
+	ErrInvalidMaintNotifications         = errors.New("hitless: invalid maintenance notifications setting (must be 'disabled', 'enabled', or 'auto')")
+	ErrMaxHandoffRetriesReached          = errors.New("hitless: max handoff retries reached")
+
+	// Configuration validation errors
+	ErrInvalidHandoffRetries = errors.New("hitless: MaxHandoffRetries must be between 1 and 10")
+)
+
+// Integration errors
+var (
+	ErrInvalidClient = errors.New("hitless: invalid client type")
+)
+
+// Handoff errors
+var (
+	ErrHandoffQueueFull = errors.New("hitless: handoff queue is full, cannot queue new handoff requests - consider increasing HandoffQueueSize or MaxWorkers in configuration")
+)
+
+// Notification errors
+var (
+	ErrInvalidNotification = errors.New("hitless: invalid notification format")
+)
+
+// connection handoff errors
+var (
+	// ErrConnectionMarkedForHandoff is returned when a connection is marked for handoff
+	// and should not be used until the handoff is complete
+	ErrConnectionMarkedForHandoff = errors.New("hitless: connection marked for handoff")
+	// ErrConnectionInvalidHandoffState is returned when a connection is in an invalid state for handoff
+	ErrConnectionInvalidHandoffState = errors.New("hitless: connection is in invalid state for handoff")
+)
+
+// general errors
+var (
+	ErrShutdown = errors.New("hitless: shutdown")
+)
+
+// circuit breaker errors
+var (
+	ErrCircuitBreakerOpen = errors.New("hitless: circuit breaker is open, failing fast")
+)
+
+// CircuitBreakerError provides detailed context for circuit breaker failures
+type CircuitBreakerError struct {
+	Endpoint     string
+	State        string
+	Failures     int64
+	LastFailure  time.Time
+	NextAttempt  time.Time
+	Message      string
+}
+
+func (e *CircuitBreakerError) Error() string {
+	if e.NextAttempt.IsZero() {
+		return fmt.Sprintf("hitless: circuit breaker %s for %s (failures: %d, last: %v): %s",
+			e.State, e.Endpoint, e.Failures, e.LastFailure, e.Message)
+	}
+	return fmt.Sprintf("hitless: circuit breaker %s for %s (failures: %d, last: %v, next attempt: %v): %s",
+		e.State, e.Endpoint, e.Failures, e.LastFailure, e.NextAttempt, e.Message)
+}
+
+// HandoffError provides detailed context for connection handoff failures
+type HandoffError struct {
+	ConnectionID   uint64
+	SourceEndpoint string
+	TargetEndpoint string
+	Attempt        int
+	MaxAttempts    int
+	Duration       time.Duration
+	FinalError     error
+	Message        string
+}
+
+func (e *HandoffError) Error() string {
+	return fmt.Sprintf("hitless: handoff failed for conn[%d] %s→%s (attempt %d/%d, duration: %v): %s",
+		e.ConnectionID, e.SourceEndpoint, e.TargetEndpoint,
+		e.Attempt, e.MaxAttempts, e.Duration, e.Message)
+}
+
+func (e *HandoffError) Unwrap() error {
+	return e.FinalError
+}
+
+// circuit breaker configuration errors
+var (
+	ErrInvalidCircuitBreakerFailureThreshold = errors.New("hitless: circuit breaker failure threshold must be >= 1")
+	ErrInvalidCircuitBreakerResetTimeout     = errors.New("hitless: circuit breaker reset timeout must be >= 0")
+	ErrInvalidCircuitBreakerMaxRequests      = errors.New("hitless: circuit breaker max requests must be >= 1")
+)

hitless/example_hooks.go

@@ -0,0 +1,100 @@
+package hitless
+
+import (
+	"context"
+	"fmt"
+	"time"
+
+	"github.com/redis/go-redis/v9/internal"
+	"github.com/redis/go-redis/v9/internal/pool"
+	"github.com/redis/go-redis/v9/push"
+)
+
+// contextKey is a custom type for context keys to avoid collisions
+type contextKey string
+
+const (
+	startTimeKey contextKey = "notif_hitless_start_time"
+)
+
+// MetricsHook collects metrics about notification processing.
+type MetricsHook struct {
+	NotificationCounts map[string]int64
+	ProcessingTimes    map[string]time.Duration
+	ErrorCounts        map[string]int64
+	HandoffCounts      int64 // Total handoffs initiated
+	HandoffSuccesses   int64 // Successful handoffs
+	HandoffFailures    int64 // Failed handoffs
+}
+
+// NewMetricsHook creates a new metrics collection hook.
+func NewMetricsHook() *MetricsHook {
+	return &MetricsHook{
+		NotificationCounts: make(map[string]int64),
+		ProcessingTimes:    make(map[string]time.Duration),
+		ErrorCounts:        make(map[string]int64),
+	}
+}
+
+// PreHook records the start time for processing metrics.
+func (mh *MetricsHook) PreHook(ctx context.Context, notificationCtx push.NotificationHandlerContext, notificationType string, notification []interface{}) ([]interface{}, bool) {
+	mh.NotificationCounts[notificationType]++
+
+	// Log connection information if available
+	if conn, ok := notificationCtx.Conn.(*pool.Conn); ok {
+		internal.Logger.Printf(ctx, "hitless: metrics hook processing %s notification on conn[%d]", notificationType, conn.GetID())
+	}
+
+	// Store start time in context for duration calculation
+	startTime := time.Now()
+	_ = context.WithValue(ctx, startTimeKey, startTime) // Context not used further
+
+	return notification, true
+}
+
+// PostHook records processing completion and any errors.
+func (mh *MetricsHook) PostHook(ctx context.Context, notificationCtx push.NotificationHandlerContext, notificationType string, notification []interface{}, result error) {
+	// Calculate processing duration
+	if startTime, ok := ctx.Value(startTimeKey).(time.Time); ok {
+		duration := time.Since(startTime)
+		mh.ProcessingTimes[notificationType] = duration
+	}
+
+	// Record errors
+	if result != nil {
+		mh.ErrorCounts[notificationType]++
+
+		// Log error details with connection information
+		if conn, ok := notificationCtx.Conn.(*pool.Conn); ok {
+			internal.Logger.Printf(ctx, "hitless: metrics hook recorded error for %s notification on conn[%d]: %v", notificationType, conn.GetID(), result)
+		}
+	}
+}
+
+// GetMetrics returns a summary of collected metrics.
+func (mh *MetricsHook) GetMetrics() map[string]interface{} {
+	return map[string]interface{}{
+		"notification_counts": mh.NotificationCounts,
+		"processing_times":    mh.ProcessingTimes,
+		"error_counts":        mh.ErrorCounts,
+	}
+}
+
+// ExampleCircuitBreakerMonitor demonstrates how to monitor circuit breaker status
+func ExampleCircuitBreakerMonitor(poolHook *PoolHook) {
+	// Get circuit breaker statistics
+	stats := poolHook.GetCircuitBreakerStats()
+
+	for _, stat := range stats {
+		fmt.Printf("Circuit Breaker for %s:\n", stat.Endpoint)
+		fmt.Printf("  State: %s\n", stat.State)
+		fmt.Printf("  Failures: %d\n", stat.Failures)
+		fmt.Printf("  Last Failure: %v\n", stat.LastFailureTime)
+		fmt.Printf("  Last Success: %v\n", stat.LastSuccessTime)
+
+		// Alert if circuit breaker is open
+		if stat.State.String() == "open" {
+			fmt.Printf("  ⚠️  ALERT: Circuit breaker is OPEN for %s\n", stat.Endpoint)
+		}
+	}
+}

hitless/handoff_worker.go

@@ -0,0 +1,468 @@
+package hitless
+
+import (
+	"context"
+	"errors"
+	"net"
+	"sync"
+	"sync/atomic"
+	"time"
+
+	"github.com/redis/go-redis/v9/internal"
+	"github.com/redis/go-redis/v9/internal/pool"
+)
+
+// handoffWorkerManager manages background workers and queue for connection handoffs
+type handoffWorkerManager struct {
+	// Event-driven handoff support
+	handoffQueue chan HandoffRequest // Queue for handoff requests
+	shutdown     chan struct{}       // Shutdown signal
+	shutdownOnce sync.Once           // Ensure clean shutdown
+	workerWg     sync.WaitGroup      // Track worker goroutines
+
+	// On-demand worker management
+	maxWorkers     int
+	activeWorkers  atomic.Int32
+	workerTimeout  time.Duration // How long workers wait for work before exiting
+	workersScaling atomic.Bool
+
+	// Simple state tracking
+	pending sync.Map // map[uint64]int64 (connID -> seqID)
+
+	// Configuration for the hitless upgrade
+	config *Config
+
+	// Pool hook reference for handoff processing
+	poolHook *PoolHook
+
+	// Circuit breaker manager for endpoint failure handling
+	circuitBreakerManager *CircuitBreakerManager
+}
+
+// newHandoffWorkerManager creates a new handoff worker manager
+func newHandoffWorkerManager(config *Config, poolHook *PoolHook) *handoffWorkerManager {
+	return &handoffWorkerManager{
+		handoffQueue:          make(chan HandoffRequest, config.HandoffQueueSize),
+		shutdown:              make(chan struct{}),
+		maxWorkers:            config.MaxWorkers,
+		activeWorkers:         atomic.Int32{},   // Start with no workers - create on demand
+		workerTimeout:         15 * time.Second, // Workers exit after 15s of inactivity
+		config:                config,
+		poolHook:              poolHook,
+		circuitBreakerManager: newCircuitBreakerManager(config),
+	}
+}
+
+// getCurrentWorkers returns the current number of active workers (for testing)
+func (hwm *handoffWorkerManager) getCurrentWorkers() int {
+	return int(hwm.activeWorkers.Load())
+}
+
+// getPendingMap returns the pending map for testing purposes
+func (hwm *handoffWorkerManager) getPendingMap() *sync.Map {
+	return &hwm.pending
+}
+
+// getMaxWorkers returns the max workers for testing purposes
+func (hwm *handoffWorkerManager) getMaxWorkers() int {
+	return hwm.maxWorkers
+}
+
+// getHandoffQueue returns the handoff queue for testing purposes
+func (hwm *handoffWorkerManager) getHandoffQueue() chan HandoffRequest {
+	return hwm.handoffQueue
+}
+
+// getCircuitBreakerStats returns circuit breaker statistics for monitoring
+func (hwm *handoffWorkerManager) getCircuitBreakerStats() []CircuitBreakerStats {
+	return hwm.circuitBreakerManager.GetAllStats()
+}
+
+// resetCircuitBreakers resets all circuit breakers (useful for testing)
+func (hwm *handoffWorkerManager) resetCircuitBreakers() {
+	hwm.circuitBreakerManager.Reset()
+}
+
+// isHandoffPending returns true if the given connection has a pending handoff
+func (hwm *handoffWorkerManager) isHandoffPending(conn *pool.Conn) bool {
+	_, pending := hwm.pending.Load(conn.GetID())
+	return pending
+}
+
+// ensureWorkerAvailable ensures at least one worker is available to process requests
+// Creates a new worker if needed and under the max limit
+func (hwm *handoffWorkerManager) ensureWorkerAvailable() {
+	select {
+	case <-hwm.shutdown:
+		return
+	default:
+		if hwm.workersScaling.CompareAndSwap(false, true) {
+			defer hwm.workersScaling.Store(false)
+			// Check if we need a new worker
+			currentWorkers := hwm.activeWorkers.Load()
+			workersWas := currentWorkers
+			for currentWorkers < int32(hwm.maxWorkers) {
+				hwm.workerWg.Add(1)
+				go hwm.onDemandWorker()
+				currentWorkers++
+			}
+			// workersWas is always <= currentWorkers
+			// currentWorkers will be maxWorkers, but if we have a worker that was closed
+			// while we were creating new workers, just add the difference between
+			// the currentWorkers and the number of workers we observed initially (i.e. the number of workers we created)
+			hwm.activeWorkers.Add(currentWorkers - workersWas)
+		}
+	}
+}
+
+// onDemandWorker processes handoff requests and exits when idle
+func (hwm *handoffWorkerManager) onDemandWorker() {
+	defer func() {
+		// Handle panics to ensure proper cleanup
+		if r := recover(); r != nil {
+			internal.Logger.Printf(context.Background(),
+				"hitless: worker panic recovered: %v", r)
+		}
+
+		// Decrement active worker count when exiting
+		hwm.activeWorkers.Add(-1)
+		hwm.workerWg.Done()
+	}()
+
+	// Create reusable timer to prevent timer leaks
+	timer := time.NewTimer(hwm.workerTimeout)
+	defer timer.Stop()
+
+	for {
+		// Reset timer for next iteration
+		if !timer.Stop() {
+			select {
+			case <-timer.C:
+			default:
+			}
+		}
+		timer.Reset(hwm.workerTimeout)
+
+		select {
+		case <-hwm.shutdown:
+			return
+		case <-timer.C:
+			// Worker has been idle for too long, exit to save resources
+			if hwm.config != nil && hwm.config.LogLevel.InfoOrAbove() {
+				internal.Logger.Printf(context.Background(),
+					"hitless: worker exiting due to inactivity timeout (%v)", hwm.workerTimeout)
+			}
+			return
+		case request := <-hwm.handoffQueue:
+			// Check for shutdown before processing
+			select {
+			case <-hwm.shutdown:
+				// Clean up the request before exiting
+				hwm.pending.Delete(request.ConnID)
+				return
+			default:
+				// Process the request
+				hwm.processHandoffRequest(request)
+			}
+		}
+	}
+}
+
+// processHandoffRequest processes a single handoff request
+func (hwm *handoffWorkerManager) processHandoffRequest(request HandoffRequest) {
+	// Remove from pending map
+	defer hwm.pending.Delete(request.Conn.GetID())
+	internal.Logger.Printf(context.Background(), "hitless: conn[%d] Processing handoff request start", request.Conn.GetID())
+
+	// Create a context with handoff timeout from config
+	handoffTimeout := 15 * time.Second // Default timeout
+	if hwm.config != nil && hwm.config.HandoffTimeout > 0 {
+		handoffTimeout = hwm.config.HandoffTimeout
+	}
+	ctx, cancel := context.WithTimeout(context.Background(), handoffTimeout)
+	defer cancel()
+
+	// Create a context that also respects the shutdown signal
+	shutdownCtx, shutdownCancel := context.WithCancel(ctx)
+	defer shutdownCancel()
+
+	// Monitor shutdown signal in a separate goroutine
+	go func() {
+		select {
+		case <-hwm.shutdown:
+			shutdownCancel()
+		case <-shutdownCtx.Done():
+		}
+	}()
+
+	// Perform the handoff with cancellable context
+	shouldRetry, err := hwm.performConnectionHandoff(shutdownCtx, request.Conn)
+	minRetryBackoff := 500 * time.Millisecond
+	if err != nil {
+		if shouldRetry {
+			now := time.Now()
+			deadline, ok := shutdownCtx.Deadline()
+			thirdOfTimeout := handoffTimeout / 3
+			if !ok || deadline.Before(now) {
+				// wait half the timeout before retrying if no deadline or deadline has passed
+				deadline = now.Add(thirdOfTimeout)
+			}
+			afterTime := deadline.Sub(now)
+			if afterTime < minRetryBackoff {
+				afterTime = minRetryBackoff
+			}
+
+			internal.Logger.Printf(context.Background(), "Handoff failed for conn[%d] WILL RETRY After %v: %v", request.ConnID, afterTime, err)
+			time.AfterFunc(afterTime, func() {
+				if err := hwm.queueHandoff(request.Conn); err != nil {
+					internal.Logger.Printf(context.Background(), "can't queue handoff for retry: %v", err)
+					hwm.closeConnFromRequest(context.Background(), request, err)
+				}
+			})
+			return
+		} else {
+			go hwm.closeConnFromRequest(ctx, request, err)
+		}
+
+		// Clear handoff state if not returned for retry
+		seqID := request.Conn.GetMovingSeqID()
+		connID := request.Conn.GetID()
+		if hwm.poolHook.hitlessManager != nil {
+			hwm.poolHook.hitlessManager.UntrackOperationWithConnID(seqID, connID)
+		}
+	}
+}
+
+// queueHandoff queues a handoff request for processing
+// if err is returned, connection will be removed from pool
+func (hwm *handoffWorkerManager) queueHandoff(conn *pool.Conn) error {
+	// Get handoff info atomically to prevent race conditions
+	shouldHandoff, endpoint, seqID := conn.GetHandoffInfo()
+	if !shouldHandoff {
+		return errors.New("connection is not marked for handoff")
+	}
+
+	// Create handoff request with atomically retrieved data
+	request := HandoffRequest{
+		Conn:     conn,
+		ConnID:   conn.GetID(),
+		Endpoint: endpoint,
+		SeqID:    seqID,
+		Pool:     hwm.poolHook.pool, // Include pool for connection removal on failure
+	}
+
+	select {
+	// priority to shutdown
+	case <-hwm.shutdown:
+		return ErrShutdown
+	default:
+		select {
+		case <-hwm.shutdown:
+			return ErrShutdown
+		case hwm.handoffQueue <- request:
+			// Store in pending map
+			hwm.pending.Store(request.ConnID, request.SeqID)
+			// Ensure we have a worker to process this request
+			hwm.ensureWorkerAvailable()
+			return nil
+		default:
+			select {
+			case <-hwm.shutdown:
+				return ErrShutdown
+			case hwm.handoffQueue <- request:
+				// Store in pending map
+				hwm.pending.Store(request.ConnID, request.SeqID)
+				// Ensure we have a worker to process this request
+				hwm.ensureWorkerAvailable()
+				return nil
+			case <-time.After(100 * time.Millisecond): // give workers a chance to process
+				// Queue is full - log and attempt scaling
+				queueLen := len(hwm.handoffQueue)
+				queueCap := cap(hwm.handoffQueue)
+				if hwm.config != nil && hwm.config.LogLevel.WarnOrAbove() { // Warning level
+					internal.Logger.Printf(context.Background(),
+						"hitless: handoff queue is full (%d/%d), cannot queue new handoff requests - consider increasing HandoffQueueSize or MaxWorkers in configuration",
+						queueLen, queueCap)
+				}
+			}
+		}
+	}
+
+	// Ensure we have workers available to handle the load
+	hwm.ensureWorkerAvailable()
+	return ErrHandoffQueueFull
+}
+
+// shutdownWorkers gracefully shuts down the worker manager, waiting for workers to complete
+func (hwm *handoffWorkerManager) shutdownWorkers(ctx context.Context) error {
+	hwm.shutdownOnce.Do(func() {
+		close(hwm.shutdown)
+		// workers will exit when they finish their current request
+
+		// Shutdown circuit breaker manager cleanup goroutine
+		if hwm.circuitBreakerManager != nil {
+			hwm.circuitBreakerManager.Shutdown()
+		}
+	})
+
+	// Wait for workers to complete
+	done := make(chan struct{})
+	go func() {
+		hwm.workerWg.Wait()
+		close(done)
+	}()
+
+	select {
+	case <-done:
+		return nil
+	case <-ctx.Done():
+		return ctx.Err()
+	}
+}
+
+// performConnectionHandoff performs the actual connection handoff
+// When error is returned, the connection handoff should be retried if err is not ErrMaxHandoffRetriesReached
+func (hwm *handoffWorkerManager) performConnectionHandoff(ctx context.Context, conn *pool.Conn) (shouldRetry bool, err error) {
+	// Clear handoff state after successful handoff
+	connID := conn.GetID()
+
+	newEndpoint := conn.GetHandoffEndpoint()
+	if newEndpoint == "" {
+		return false, ErrConnectionInvalidHandoffState
+	}
+
+	// Use circuit breaker to protect against failing endpoints
+	circuitBreaker := hwm.circuitBreakerManager.GetCircuitBreaker(newEndpoint)
+
+	// Check if circuit breaker is open before attempting handoff
+	if circuitBreaker.IsOpen() {
+		internal.Logger.Printf(ctx, "hitless: conn[%d] handoff to %s failed fast due to circuit breaker", connID, newEndpoint)
+		return false, ErrCircuitBreakerOpen // Don't retry when circuit breaker is open
+	}
+
+	// Perform the handoff
+	shouldRetry, err = hwm.performHandoffInternal(ctx, conn, newEndpoint, connID)
+
+	// Update circuit breaker based on result
+	if err != nil {
+		// Only track dial/network errors in circuit breaker, not initialization errors
+		if shouldRetry {
+			circuitBreaker.recordFailure()
+		}
+		return shouldRetry, err
+	}
+
+	// Success - record in circuit breaker
+	circuitBreaker.recordSuccess()
+	return false, nil
+}
+
+// performHandoffInternal performs the actual handoff logic (extracted for circuit breaker integration)
+func (hwm *handoffWorkerManager) performHandoffInternal(ctx context.Context, conn *pool.Conn, newEndpoint string, connID uint64) (shouldRetry bool, err error) {
+
+	retries := conn.IncrementAndGetHandoffRetries(1)
+	internal.Logger.Printf(ctx, "hitless: conn[%d] Retry %d: Performing handoff to %s(was %s)", connID, retries, newEndpoint, conn.RemoteAddr().String())
+	maxRetries := 3 // Default fallback
+	if hwm.config != nil {
+		maxRetries = hwm.config.MaxHandoffRetries
+	}
+
+	if retries > maxRetries {
+		if hwm.config != nil && hwm.config.LogLevel.WarnOrAbove() { // Warning level
+			internal.Logger.Printf(ctx,
+				"hitless: reached max retries (%d) for handoff of conn[%d] to %s",
+				maxRetries, connID, newEndpoint)
+		}
+		// won't retry on ErrMaxHandoffRetriesReached
+		return false, ErrMaxHandoffRetriesReached
+	}
+
+	// Create endpoint-specific dialer
+	endpointDialer := hwm.createEndpointDialer(newEndpoint)
+
+	// Create new connection to the new endpoint
+	newNetConn, err := endpointDialer(ctx)
+	if err != nil {
+		internal.Logger.Printf(ctx, "hitless: conn[%d] Failed to dial new endpoint %s: %v", connID, newEndpoint, err)
+		// hitless: will retry
+		// Maybe a network error - retry after a delay
+		return true, err
+	}
+
+	// Get the old connection
+	oldConn := conn.GetNetConn()
+
+	// Apply relaxed timeout to the new connection for the configured post-handoff duration
+	// This gives the new connection more time to handle operations during cluster transition
+	// Setting this here (before initing the connection) ensures that the connection is going
+	// to use the relaxed timeout for the first operation (auth/ACL select)
+	if hwm.config != nil && hwm.config.PostHandoffRelaxedDuration > 0 {
+		relaxedTimeout := hwm.config.RelaxedTimeout
+		// Set relaxed timeout with deadline - no background goroutine needed
+		deadline := time.Now().Add(hwm.config.PostHandoffRelaxedDuration)
+		conn.SetRelaxedTimeoutWithDeadline(relaxedTimeout, relaxedTimeout, deadline)
+
+		if hwm.config.LogLevel.InfoOrAbove() {
+			internal.Logger.Printf(context.Background(),
+				"hitless: conn[%d] applied post-handoff relaxed timeout (%v) until %v",
+				connID, relaxedTimeout, deadline.Format("15:04:05.000"))
+		}
+	}
+
+	// Replace the connection and execute initialization
+	err = conn.SetNetConnAndInitConn(ctx, newNetConn)
+	if err != nil {
+		// hitless: won't retry
+		// Initialization failed - remove the connection
+		return false, err
+	}
+	defer func() {
+		if oldConn != nil {
+			oldConn.Close()
+		}
+	}()
+
+	conn.ClearHandoffState()
+	internal.Logger.Printf(ctx, "hitless: conn[%d] Handoff to %s successful", connID, newEndpoint)
+
+	return false, nil
+}
+
+// createEndpointDialer creates a dialer function that connects to a specific endpoint
+func (hwm *handoffWorkerManager) createEndpointDialer(endpoint string) func(context.Context) (net.Conn, error) {
+	return func(ctx context.Context) (net.Conn, error) {
+		// Parse endpoint to extract host and port
+		host, port, err := net.SplitHostPort(endpoint)
+		if err != nil {
+			// If no port specified, assume default Redis port
+			host = endpoint
+			if port == "" {
+				port = "6379"
+			}
+		}
+
+		// Use the base dialer to connect to the new endpoint
+		return hwm.poolHook.baseDialer(ctx, hwm.poolHook.network, net.JoinHostPort(host, port))
+	}
+}
+
+// closeConnFromRequest closes the connection and logs the reason
+func (hwm *handoffWorkerManager) closeConnFromRequest(ctx context.Context, request HandoffRequest, err error) {
+	pooler := request.Pool
+	conn := request.Conn
+	if pooler != nil {
+		pooler.Remove(ctx, conn, err)
+		if hwm.config != nil && hwm.config.LogLevel.WarnOrAbove() { // Warning level
+			internal.Logger.Printf(ctx,
+				"hitless: removed conn[%d] from pool due: %v",
+				conn.GetID(), err)
+		}
+	} else {
+		conn.Close()
+		if hwm.config != nil && hwm.config.LogLevel.WarnOrAbove() { // Warning level
+			internal.Logger.Printf(ctx,
+				"hitless: no pool provided for conn[%d], cannot remove due to: %v",
+				conn.GetID(), err)
+		}
+	}
+}

hitless/hitless_manager.go

@@ -0,0 +1,318 @@
+package hitless
+
+import (
+	"context"
+	"fmt"
+	"net"
+	"sync"
+	"sync/atomic"
+	"time"
+
+	"github.com/redis/go-redis/v9/internal"
+	"github.com/redis/go-redis/v9/internal/interfaces"
+	"github.com/redis/go-redis/v9/internal/pool"
+	"github.com/redis/go-redis/v9/push"
+)
+
+// Push notification type constants for hitless upgrades
+const (
+	NotificationMoving      = "MOVING"
+	NotificationMigrating   = "MIGRATING"
+	NotificationMigrated    = "MIGRATED"
+	NotificationFailingOver = "FAILING_OVER"
+	NotificationFailedOver  = "FAILED_OVER"
+)
+
+// hitlessNotificationTypes contains all notification types that hitless upgrades handles
+var hitlessNotificationTypes = []string{
+	NotificationMoving,
+	NotificationMigrating,
+	NotificationMigrated,
+	NotificationFailingOver,
+	NotificationFailedOver,
+}
+
+// NotificationHook is called before and after notification processing
+// PreHook can modify the notification and return false to skip processing
+// PostHook is called after successful processing
+type NotificationHook interface {
+	PreHook(ctx context.Context, notificationCtx push.NotificationHandlerContext, notificationType string, notification []interface{}) ([]interface{}, bool)
+	PostHook(ctx context.Context, notificationCtx push.NotificationHandlerContext, notificationType string, notification []interface{}, result error)
+}
+
+// MovingOperationKey provides a unique key for tracking MOVING operations
+// that combines sequence ID with connection identifier to handle duplicate
+// sequence IDs across multiple connections to the same node.
+type MovingOperationKey struct {
+	SeqID  int64  // Sequence ID from MOVING notification
+	ConnID uint64 // Unique connection identifier
+}
+
+// String returns a string representation of the key for debugging
+func (k MovingOperationKey) String() string {
+	return fmt.Sprintf("seq:%d-conn:%d", k.SeqID, k.ConnID)
+}
+
+// HitlessManager provides a simplified hitless upgrade functionality with hooks and atomic state.
+type HitlessManager struct {
+	client  interfaces.ClientInterface
+	config  *Config
+	options interfaces.OptionsInterface
+	pool    pool.Pooler
+
+	// MOVING operation tracking - using sync.Map for better concurrent performance
+	activeMovingOps sync.Map // map[MovingOperationKey]*MovingOperation
+
+	// Atomic state tracking - no locks needed for state queries
+	activeOperationCount atomic.Int64 // Number of active operations
+	closed               atomic.Bool  // Manager closed state
+
+	// Notification hooks for extensibility
+	hooks        []NotificationHook
+	hooksMu      sync.RWMutex // Protects hooks slice
+	poolHooksRef *PoolHook
+}
+
+// MovingOperation tracks an active MOVING operation.
+type MovingOperation struct {
+	SeqID       int64
+	NewEndpoint string
+	StartTime   time.Time
+	Deadline    time.Time
+}
+
+// NewHitlessManager creates a new simplified hitless manager.
+func NewHitlessManager(client interfaces.ClientInterface, pool pool.Pooler, config *Config) (*HitlessManager, error) {
+	if client == nil {
+		return nil, ErrInvalidClient
+	}
+
+	hm := &HitlessManager{
+		client:  client,
+		pool:    pool,
+		options: client.GetOptions(),
+		config:  config.Clone(),
+		hooks:   make([]NotificationHook, 0),
+	}
+
+	// Set up push notification handling
+	if err := hm.setupPushNotifications(); err != nil {
+		return nil, err
+	}
+
+	return hm, nil
+}
+
+// GetPoolHook creates a pool hook with a custom dialer.
+func (hm *HitlessManager) InitPoolHook(baseDialer func(context.Context, string, string) (net.Conn, error)) {
+	poolHook := hm.createPoolHook(baseDialer)
+	hm.pool.AddPoolHook(poolHook)
+}
+
+// setupPushNotifications sets up push notification handling by registering with the client's processor.
+func (hm *HitlessManager) setupPushNotifications() error {
+	processor := hm.client.GetPushProcessor()
+	if processor == nil {
+		return ErrInvalidClient // Client doesn't support push notifications
+	}
+
+	// Create our notification handler
+	handler := &NotificationHandler{manager: hm}
+
+	// Register handlers for all hitless upgrade notifications with the client's processor
+	for _, notificationType := range hitlessNotificationTypes {
+		if err := processor.RegisterHandler(notificationType, handler, true); err != nil {
+			return fmt.Errorf("failed to register handler for %s: %w", notificationType, err)
+		}
+	}
+
+	return nil
+}
+
+// TrackMovingOperationWithConnID starts a new MOVING operation with a specific connection ID.
+func (hm *HitlessManager) TrackMovingOperationWithConnID(ctx context.Context, newEndpoint string, deadline time.Time, seqID int64, connID uint64) error {
+	// Create composite key
+	key := MovingOperationKey{
+		SeqID:  seqID,
+		ConnID: connID,
+	}
+
+	// Create MOVING operation record
+	movingOp := &MovingOperation{
+		SeqID:       seqID,
+		NewEndpoint: newEndpoint,
+		StartTime:   time.Now(),
+		Deadline:    deadline,
+	}
+
+	// Use LoadOrStore for atomic check-and-set operation
+	if _, loaded := hm.activeMovingOps.LoadOrStore(key, movingOp); loaded {
+		// Duplicate MOVING notification, ignore
+		if hm.config.LogLevel.DebugOrAbove() { // Debug level
+			internal.Logger.Printf(context.Background(), "hitless: conn[%d] seqID[%d] Duplicate MOVING operation ignored: %s", connID, seqID, key.String())
+		}
+		return nil
+	}
+	if hm.config.LogLevel.DebugOrAbove() { // Debug level
+		internal.Logger.Printf(context.Background(), "hitless: conn[%d] seqID[%d] Tracking MOVING operation: %s", connID, seqID, key.String())
+	}
+
+	// Increment active operation count atomically
+	hm.activeOperationCount.Add(1)
+
+	return nil
+}
+
+// UntrackOperationWithConnID completes a MOVING operation with a specific connection ID.
+func (hm *HitlessManager) UntrackOperationWithConnID(seqID int64, connID uint64) {
+	// Create composite key
+	key := MovingOperationKey{
+		SeqID:  seqID,
+		ConnID: connID,
+	}
+
+	// Remove from active operations atomically
+	if _, loaded := hm.activeMovingOps.LoadAndDelete(key); loaded {
+		if hm.config.LogLevel.DebugOrAbove() { // Debug level
+			internal.Logger.Printf(context.Background(), "hitless: conn[%d] seqID[%d] Untracking MOVING operation: %s", connID, seqID, key.String())
+		}
+		// Decrement active operation count only if operation existed
+		hm.activeOperationCount.Add(-1)
+	} else {
+		if hm.config.LogLevel.DebugOrAbove() { // Debug level
+			internal.Logger.Printf(context.Background(), "hitless: conn[%d] seqID[%d] Operation not found for untracking: %s", connID, seqID, key.String())
+		}
+	}
+}
+
+// GetActiveMovingOperations returns active operations with composite keys.
+// WARNING: This method creates a new map and copies all operations on every call.
+// Use sparingly, especially in hot paths or high-frequency logging.
+func (hm *HitlessManager) GetActiveMovingOperations() map[MovingOperationKey]*MovingOperation {
+	result := make(map[MovingOperationKey]*MovingOperation)
+
+	// Iterate over sync.Map to build result
+	hm.activeMovingOps.Range(func(key, value interface{}) bool {
+		k := key.(MovingOperationKey)
+		op := value.(*MovingOperation)
+
+		// Create a copy to avoid sharing references
+		result[k] = &MovingOperation{
+			SeqID:       op.SeqID,
+			NewEndpoint: op.NewEndpoint,
+			StartTime:   op.StartTime,
+			Deadline:    op.Deadline,
+		}
+		return true // Continue iteration
+	})
+
+	return result
+}
+
+// IsHandoffInProgress returns true if any handoff is in progress.
+// Uses atomic counter for lock-free operation.
+func (hm *HitlessManager) IsHandoffInProgress() bool {
+	return hm.activeOperationCount.Load() > 0
+}
+
+// GetActiveOperationCount returns the number of active operations.
+// Uses atomic counter for lock-free operation.
+func (hm *HitlessManager) GetActiveOperationCount() int64 {
+	return hm.activeOperationCount.Load()
+}
+
+// Close closes the hitless manager.
+func (hm *HitlessManager) Close() error {
+	// Use atomic operation for thread-safe close check
+	if !hm.closed.CompareAndSwap(false, true) {
+		return nil // Already closed
+	}
+
+	// Shutdown the pool hook if it exists
+	if hm.poolHooksRef != nil {
+		// Use a timeout to prevent hanging indefinitely
+		shutdownCtx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+		defer cancel()
+
+		err := hm.poolHooksRef.Shutdown(shutdownCtx)
+		if err != nil {
+			// was not able to close pool hook, keep closed state false
+			hm.closed.Store(false)
+			return err
+		}
+		// Remove the pool hook from the pool
+		if hm.pool != nil {
+			hm.pool.RemovePoolHook(hm.poolHooksRef)
+		}
+	}
+
+	// Clear all active operations
+	hm.activeMovingOps.Range(func(key, value interface{}) bool {
+		hm.activeMovingOps.Delete(key)
+		return true
+	})
+
+	// Reset counter
+	hm.activeOperationCount.Store(0)
+
+	return nil
+}
+
+// GetState returns current state using atomic counter for lock-free operation.
+func (hm *HitlessManager) GetState() State {
+	if hm.activeOperationCount.Load() > 0 {
+		return StateMoving
+	}
+	return StateIdle
+}
+
+// processPreHooks calls all pre-hooks and returns the modified notification and whether to continue processing.
+func (hm *HitlessManager) processPreHooks(ctx context.Context, notificationCtx push.NotificationHandlerContext, notificationType string, notification []interface{}) ([]interface{}, bool) {
+	hm.hooksMu.RLock()
+	defer hm.hooksMu.RUnlock()
+
+	currentNotification := notification
+
+	for _, hook := range hm.hooks {
+		modifiedNotification, shouldContinue := hook.PreHook(ctx, notificationCtx, notificationType, currentNotification)
+		if !shouldContinue {
+			return modifiedNotification, false
+		}
+		currentNotification = modifiedNotification
+	}
+
+	return currentNotification, true
+}
+
+// processPostHooks calls all post-hooks with the processing result.
+func (hm *HitlessManager) processPostHooks(ctx context.Context, notificationCtx push.NotificationHandlerContext, notificationType string, notification []interface{}, result error) {
+	hm.hooksMu.RLock()
+	defer hm.hooksMu.RUnlock()
+
+	for _, hook := range hm.hooks {
+		hook.PostHook(ctx, notificationCtx, notificationType, notification, result)
+	}
+}
+
+// createPoolHook creates a pool hook with this manager already set.
+func (hm *HitlessManager) createPoolHook(baseDialer func(context.Context, string, string) (net.Conn, error)) *PoolHook {
+	if hm.poolHooksRef != nil {
+		return hm.poolHooksRef
+	}
+	// Get pool size from client options for better worker defaults
+	poolSize := 0
+	if hm.options != nil {
+		poolSize = hm.options.GetPoolSize()
+	}
+
+	hm.poolHooksRef = NewPoolHookWithPoolSize(baseDialer, hm.options.GetNetwork(), hm.config, hm, poolSize)
+	hm.poolHooksRef.SetPool(hm.pool)
+
+	return hm.poolHooksRef
+}
+
+func (hm *HitlessManager) AddNotificationHook(notificationHook NotificationHook) {
+	hm.hooksMu.Lock()
+	defer hm.hooksMu.Unlock()
+	hm.hooks = append(hm.hooks, notificationHook)
+}

hitless/hitless_manager_test.go

@@ -0,0 +1,260 @@
+package hitless
+
+import (
+	"context"
+	"net"
+	"testing"
+	"time"
+
+	"github.com/redis/go-redis/v9/internal/interfaces"
+)
+
+// MockClient implements interfaces.ClientInterface for testing
+type MockClient struct {
+	options interfaces.OptionsInterface
+}
+
+func (mc *MockClient) GetOptions() interfaces.OptionsInterface {
+	return mc.options
+}
+
+func (mc *MockClient) GetPushProcessor() interfaces.NotificationProcessor {
+	return &MockPushProcessor{}
+}
+
+// MockPushProcessor implements interfaces.NotificationProcessor for testing
+type MockPushProcessor struct{}
+
+func (mpp *MockPushProcessor) RegisterHandler(notificationType string, handler interface{}, protected bool) error {
+	return nil
+}
+
+func (mpp *MockPushProcessor) UnregisterHandler(pushNotificationName string) error {
+	return nil
+}
+
+func (mpp *MockPushProcessor) GetHandler(pushNotificationName string) interface{} {
+	return nil
+}
+
+// MockOptions implements interfaces.OptionsInterface for testing
+type MockOptions struct{}
+
+func (mo *MockOptions) GetReadTimeout() time.Duration {
+	return 5 * time.Second
+}
+
+func (mo *MockOptions) GetWriteTimeout() time.Duration {
+	return 5 * time.Second
+}
+
+func (mo *MockOptions) GetAddr() string {
+	return "localhost:6379"
+}
+
+func (mo *MockOptions) IsTLSEnabled() bool {
+	return false
+}
+
+func (mo *MockOptions) GetProtocol() int {
+	return 3 // RESP3
+}
+
+func (mo *MockOptions) GetPoolSize() int {
+	return 10
+}
+
+func (mo *MockOptions) GetNetwork() string {
+	return "tcp"
+}
+
+func (mo *MockOptions) NewDialer() func(context.Context) (net.Conn, error) {
+	return func(ctx context.Context) (net.Conn, error) {
+		return nil, nil
+	}
+}
+
+func TestHitlessManagerRefactoring(t *testing.T) {
+	t.Run("AtomicStateTracking", func(t *testing.T) {
+		config := DefaultConfig()
+		client := &MockClient{options: &MockOptions{}}
+
+		manager, err := NewHitlessManager(client, nil, config)
+		if err != nil {
+			t.Fatalf("Failed to create hitless manager: %v", err)
+		}
+		defer manager.Close()
+
+		// Test initial state
+		if manager.IsHandoffInProgress() {
+			t.Error("Expected no handoff in progress initially")
+		}
+
+		if manager.GetActiveOperationCount() != 0 {
+			t.Errorf("Expected 0 active operations, got %d", manager.GetActiveOperationCount())
+		}
+
+		if manager.GetState() != StateIdle {
+			t.Errorf("Expected StateIdle, got %v", manager.GetState())
+		}
+
+		// Add an operation
+		ctx := context.Background()
+		deadline := time.Now().Add(30 * time.Second)
+		err = manager.TrackMovingOperationWithConnID(ctx, "new-endpoint:6379", deadline, 12345, 1)
+		if err != nil {
+			t.Fatalf("Failed to track operation: %v", err)
+		}
+
+		// Test state after adding operation
+		if !manager.IsHandoffInProgress() {
+			t.Error("Expected handoff in progress after adding operation")
+		}
+
+		if manager.GetActiveOperationCount() != 1 {
+			t.Errorf("Expected 1 active operation, got %d", manager.GetActiveOperationCount())
+		}
+
+		if manager.GetState() != StateMoving {
+			t.Errorf("Expected StateMoving, got %v", manager.GetState())
+		}
+
+		// Remove the operation
+		manager.UntrackOperationWithConnID(12345, 1)
+
+		// Test state after removing operation
+		if manager.IsHandoffInProgress() {
+			t.Error("Expected no handoff in progress after removing operation")
+		}
+
+		if manager.GetActiveOperationCount() != 0 {
+			t.Errorf("Expected 0 active operations, got %d", manager.GetActiveOperationCount())
+		}
+
+		if manager.GetState() != StateIdle {
+			t.Errorf("Expected StateIdle, got %v", manager.GetState())
+		}
+	})
+
+	t.Run("SyncMapPerformance", func(t *testing.T) {
+		config := DefaultConfig()
+		client := &MockClient{options: &MockOptions{}}
+
+		manager, err := NewHitlessManager(client, nil, config)
+		if err != nil {
+			t.Fatalf("Failed to create hitless manager: %v", err)
+		}
+		defer manager.Close()
+
+		ctx := context.Background()
+		deadline := time.Now().Add(30 * time.Second)
+
+		// Test concurrent operations
+		const numOps = 100
+		for i := 0; i < numOps; i++ {
+			err := manager.TrackMovingOperationWithConnID(ctx, "endpoint:6379", deadline, int64(i), uint64(i))
+			if err != nil {
+				t.Fatalf("Failed to track operation %d: %v", i, err)
+			}
+		}
+
+		if manager.GetActiveOperationCount() != numOps {
+			t.Errorf("Expected %d active operations, got %d", numOps, manager.GetActiveOperationCount())
+		}
+
+		// Test GetActiveMovingOperations
+		operations := manager.GetActiveMovingOperations()
+		if len(operations) != numOps {
+			t.Errorf("Expected %d operations in map, got %d", numOps, len(operations))
+		}
+
+		// Remove all operations
+		for i := 0; i < numOps; i++ {
+			manager.UntrackOperationWithConnID(int64(i), uint64(i))
+		}
+
+		if manager.GetActiveOperationCount() != 0 {
+			t.Errorf("Expected 0 active operations after cleanup, got %d", manager.GetActiveOperationCount())
+		}
+	})
+
+	t.Run("DuplicateOperationHandling", func(t *testing.T) {
+		config := DefaultConfig()
+		client := &MockClient{options: &MockOptions{}}
+
+		manager, err := NewHitlessManager(client, nil, config)
+		if err != nil {
+			t.Fatalf("Failed to create hitless manager: %v", err)
+		}
+		defer manager.Close()
+
+		ctx := context.Background()
+		deadline := time.Now().Add(30 * time.Second)
+
+		// Add operation
+		err = manager.TrackMovingOperationWithConnID(ctx, "endpoint:6379", deadline, 12345, 1)
+		if err != nil {
+			t.Fatalf("Failed to track operation: %v", err)
+		}
+
+		// Try to add duplicate operation
+		err = manager.TrackMovingOperationWithConnID(ctx, "endpoint:6379", deadline, 12345, 1)
+		if err != nil {
+			t.Fatalf("Duplicate operation should not return error: %v", err)
+		}
+
+		// Should still have only 1 operation
+		if manager.GetActiveOperationCount() != 1 {
+			t.Errorf("Expected 1 active operation after duplicate, got %d", manager.GetActiveOperationCount())
+		}
+	})
+
+	t.Run("NotificationTypeConstants", func(t *testing.T) {
+		// Test that constants are properly defined
+		expectedTypes := []string{
+			NotificationMoving,
+			NotificationMigrating,
+			NotificationMigrated,
+			NotificationFailingOver,
+			NotificationFailedOver,
+		}
+
+		if len(hitlessNotificationTypes) != len(expectedTypes) {
+			t.Errorf("Expected %d notification types, got %d", len(expectedTypes), len(hitlessNotificationTypes))
+		}
+
+		// Test that all expected types are present
+		typeMap := make(map[string]bool)
+		for _, t := range hitlessNotificationTypes {
+			typeMap[t] = true
+		}
+
+		for _, expected := range expectedTypes {
+			if !typeMap[expected] {
+				t.Errorf("Expected notification type %s not found in hitlessNotificationTypes", expected)
+			}
+		}
+
+		// Test that hitlessNotificationTypes contains all expected constants
+		expectedConstants := []string{
+			NotificationMoving,
+			NotificationMigrating,
+			NotificationMigrated,
+			NotificationFailingOver,
+			NotificationFailedOver,
+		}
+
+		for _, expected := range expectedConstants {
+			found := false
+			for _, actual := range hitlessNotificationTypes {
+				if actual == expected {
+					found = true
+					break
+				}
+			}
+			if !found {
+				t.Errorf("Expected constant %s not found in hitlessNotificationTypes", expected)
+			}
+		}
+	})
+}

hitless/hooks.go

@@ -0,0 +1,47 @@
+package hitless
+
+import (
+	"context"
+
+	"github.com/redis/go-redis/v9/internal"
+	"github.com/redis/go-redis/v9/internal/pool"
+	"github.com/redis/go-redis/v9/logging"
+	"github.com/redis/go-redis/v9/push"
+)
+
+// LoggingHook is an example hook implementation that logs all notifications.
+type LoggingHook struct {
+	LogLevel logging.LogLevel
+}
+
+// PreHook logs the notification before processing and allows modification.
+func (lh *LoggingHook) PreHook(ctx context.Context, notificationCtx push.NotificationHandlerContext, notificationType string, notification []interface{}) ([]interface{}, bool) {
+	if lh.LogLevel.InfoOrAbove() { // Info level
+		// Log the notification type and content
+		connID := uint64(0)
+		if conn, ok := notificationCtx.Conn.(*pool.Conn); ok {
+			connID = conn.GetID()
+		}
+		internal.Logger.Printf(ctx, "hitless: conn[%d] processing %s notification: %v", connID, notificationType, notification)
+	}
+	return notification, true // Continue processing with unmodified notification
+}
+
+// PostHook logs the result after processing.
+func (lh *LoggingHook) PostHook(ctx context.Context, notificationCtx push.NotificationHandlerContext, notificationType string, notification []interface{}, result error) {
+	connID := uint64(0)
+	if conn, ok := notificationCtx.Conn.(*pool.Conn); ok {
+		connID = conn.GetID()
+	}
+	if result != nil && lh.LogLevel.WarnOrAbove() { // Warning level
+		internal.Logger.Printf(ctx, "hitless: conn[%d] %s notification processing failed: %v - %v", connID, notificationType, result, notification)
+	} else if lh.LogLevel.DebugOrAbove() { // Debug level
+		internal.Logger.Printf(ctx, "hitless: conn[%d] %s notification processed successfully", connID, notificationType)
+	}
+}
+
+// NewLoggingHook creates a new logging hook with the specified log level.
+// Log levels: LogLevelError=errors, LogLevelWarn=warnings, LogLevelInfo=info, LogLevelDebug=debug
+func NewLoggingHook(logLevel logging.LogLevel) *LoggingHook {
+	return &LoggingHook{LogLevel: logLevel}
+}

hitless/pool_hook.go

@@ -0,0 +1,179 @@
+package hitless
+
+import (
+	"context"
+	"net"
+	"sync"
+	"time"
+
+	"github.com/redis/go-redis/v9/internal"
+	"github.com/redis/go-redis/v9/internal/pool"
+)
+
+// HitlessManagerInterface defines the interface for completing handoff operations
+type HitlessManagerInterface interface {
+	TrackMovingOperationWithConnID(ctx context.Context, newEndpoint string, deadline time.Time, seqID int64, connID uint64) error
+	UntrackOperationWithConnID(seqID int64, connID uint64)
+}
+
+// HandoffRequest represents a request to handoff a connection to a new endpoint
+type HandoffRequest struct {
+	Conn     *pool.Conn
+	ConnID   uint64 // Unique connection identifier
+	Endpoint string
+	SeqID    int64
+	Pool     pool.Pooler // Pool to remove connection from on failure
+}
+
+// PoolHook implements pool.PoolHook for Redis-specific connection handling
+// with hitless upgrade support.
+type PoolHook struct {
+	// Base dialer for creating connections to new endpoints during handoffs
+	// args are network and address
+	baseDialer func(context.Context, string, string) (net.Conn, error)
+
+	// Network type (e.g., "tcp", "unix")
+	network string
+
+	// Worker manager for background handoff processing
+	workerManager *handoffWorkerManager
+
+	// Configuration for the hitless upgrade
+	config *Config
+
+	// Hitless manager for operation completion tracking
+	hitlessManager HitlessManagerInterface
+
+	// Pool interface for removing connections on handoff failure
+	pool pool.Pooler
+}
+
+// NewPoolHook creates a new pool hook
+func NewPoolHook(baseDialer func(context.Context, string, string) (net.Conn, error), network string, config *Config, hitlessManager HitlessManagerInterface) *PoolHook {
+	return NewPoolHookWithPoolSize(baseDialer, network, config, hitlessManager, 0)
+}
+
+// NewPoolHookWithPoolSize creates a new pool hook with pool size for better worker defaults
+func NewPoolHookWithPoolSize(baseDialer func(context.Context, string, string) (net.Conn, error), network string, config *Config, hitlessManager HitlessManagerInterface, poolSize int) *PoolHook {
+	// Apply defaults if config is nil or has zero values
+	if config == nil {
+		config = config.ApplyDefaultsWithPoolSize(poolSize)
+	}
+
+	ph := &PoolHook{
+		// baseDialer is used to create connections to new endpoints during handoffs
+		baseDialer: baseDialer,
+		network:    network,
+		config:     config,
+		// Hitless manager for operation completion tracking
+		hitlessManager: hitlessManager,
+	}
+
+	// Create worker manager
+	ph.workerManager = newHandoffWorkerManager(config, ph)
+
+	return ph
+}
+
+// SetPool sets the pool interface for removing connections on handoff failure
+func (ph *PoolHook) SetPool(pooler pool.Pooler) {
+	ph.pool = pooler
+}
+
+// GetCurrentWorkers returns the current number of active workers (for testing)
+func (ph *PoolHook) GetCurrentWorkers() int {
+	return ph.workerManager.getCurrentWorkers()
+}
+
+// IsHandoffPending returns true if the given connection has a pending handoff
+func (ph *PoolHook) IsHandoffPending(conn *pool.Conn) bool {
+	return ph.workerManager.isHandoffPending(conn)
+}
+
+// GetPendingMap returns the pending map for testing purposes
+func (ph *PoolHook) GetPendingMap() *sync.Map {
+	return ph.workerManager.getPendingMap()
+}
+
+// GetMaxWorkers returns the max workers for testing purposes
+func (ph *PoolHook) GetMaxWorkers() int {
+	return ph.workerManager.getMaxWorkers()
+}
+
+// GetHandoffQueue returns the handoff queue for testing purposes
+func (ph *PoolHook) GetHandoffQueue() chan HandoffRequest {
+	return ph.workerManager.getHandoffQueue()
+}
+
+// GetCircuitBreakerStats returns circuit breaker statistics for monitoring
+func (ph *PoolHook) GetCircuitBreakerStats() []CircuitBreakerStats {
+	return ph.workerManager.getCircuitBreakerStats()
+}
+
+// ResetCircuitBreakers resets all circuit breakers (useful for testing)
+func (ph *PoolHook) ResetCircuitBreakers() {
+	ph.workerManager.resetCircuitBreakers()
+}
+
+// OnGet is called when a connection is retrieved from the pool
+func (ph *PoolHook) OnGet(ctx context.Context, conn *pool.Conn, _ bool) error {
+	// NOTE: There are two conditions to make sure we don't return a connection that should be handed off or is
+	// in a handoff state at the moment.
+
+	// Check if connection is usable (not in a handoff state)
+	// Should not happen since the pool will not return a connection that is not usable.
+	if !conn.IsUsable() {
+		return ErrConnectionMarkedForHandoff
+	}
+
+	// Check if connection is marked for handoff, which means it will be queued for handoff on put.
+	if conn.ShouldHandoff() {
+		return ErrConnectionMarkedForHandoff
+	}
+
+	return nil
+}
+
+// OnPut is called when a connection is returned to the pool
+func (ph *PoolHook) OnPut(ctx context.Context, conn *pool.Conn) (shouldPool bool, shouldRemove bool, err error) {
+	// first check if we should handoff for faster rejection
+	if !conn.ShouldHandoff() {
+		// Default behavior (no handoff): pool the connection
+		return true, false, nil
+	}
+
+	// check pending handoff to not queue the same connection twice
+	if ph.workerManager.isHandoffPending(conn) {
+		// Default behavior (pending handoff): pool the connection
+		return true, false, nil
+	}
+
+	if err := ph.workerManager.queueHandoff(conn); err != nil {
+		// Failed to queue handoff, remove the connection
+		internal.Logger.Printf(ctx, "Failed to queue handoff: %v", err)
+		// Don't pool, remove connection, no error to caller
+		return false, true, nil
+	}
+
+	// Check if handoff was already processed by a worker before we can mark it as queued
+	if !conn.ShouldHandoff() {
+		// Handoff was already processed - this is normal and the connection should be pooled
+		return true, false, nil
+	}
+
+	if err := conn.MarkQueuedForHandoff(); err != nil {
+		// If marking fails, check if handoff was processed in the meantime
+		if !conn.ShouldHandoff() {
+			// Handoff was processed - this is normal, pool the connection
+			return true, false, nil
+		}
+		// Other error - remove the connection
+		return false, true, nil
+	}
+	return true, false, nil
+}
+
+// Shutdown gracefully shuts down the processor, waiting for workers to complete
+func (ph *PoolHook) Shutdown(ctx context.Context) error {
+	return ph.workerManager.shutdownWorkers(ctx)
+}

hitless/pool_hook_test.go

@@ -0,0 +1,964 @@
+package hitless
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"net"
+	"sync"
+	"sync/atomic"
+	"testing"
+	"time"
+
+	"github.com/redis/go-redis/v9/internal/pool"
+)
+
+// mockNetConn implements net.Conn for testing
+type mockNetConn struct {
+	addr           string
+	shouldFailInit bool
+}
+
+func (m *mockNetConn) Read(b []byte) (n int, err error)   { return 0, nil }
+func (m *mockNetConn) Write(b []byte) (n int, err error)  { return len(b), nil }
+func (m *mockNetConn) Close() error                       { return nil }
+func (m *mockNetConn) LocalAddr() net.Addr                { return &mockAddr{m.addr} }
+func (m *mockNetConn) RemoteAddr() net.Addr               { return &mockAddr{m.addr} }
+func (m *mockNetConn) SetDeadline(t time.Time) error      { return nil }
+func (m *mockNetConn) SetReadDeadline(t time.Time) error  { return nil }
+func (m *mockNetConn) SetWriteDeadline(t time.Time) error { return nil }
+
+type mockAddr struct {
+	addr string
+}
+
+func (m *mockAddr) Network() string { return "tcp" }
+func (m *mockAddr) String() string  { return m.addr }
+
+// createMockPoolConnection creates a mock pool connection for testing
+func createMockPoolConnection() *pool.Conn {
+	mockNetConn := &mockNetConn{addr: "test:6379"}
+	conn := pool.NewConn(mockNetConn)
+	conn.SetUsable(true) // Make connection usable for testing
+	return conn
+}
+
+// mockPool implements pool.Pooler for testing
+type mockPool struct {
+	removedConnections map[uint64]bool
+	mu                 sync.Mutex
+}
+
+func (mp *mockPool) NewConn(ctx context.Context) (*pool.Conn, error) {
+	return nil, errors.New("not implemented")
+}
+
+func (mp *mockPool) CloseConn(conn *pool.Conn) error {
+	return nil
+}
+
+func (mp *mockPool) Get(ctx context.Context) (*pool.Conn, error) {
+	return nil, errors.New("not implemented")
+}
+
+func (mp *mockPool) Put(ctx context.Context, conn *pool.Conn) {
+	// Not implemented for testing
+}
+
+func (mp *mockPool) Remove(ctx context.Context, conn *pool.Conn, reason error) {
+	mp.mu.Lock()
+	defer mp.mu.Unlock()
+
+	// Use pool.Conn directly - no adapter needed
+	mp.removedConnections[conn.GetID()] = true
+}
+
+// WasRemoved safely checks if a connection was removed from the pool
+func (mp *mockPool) WasRemoved(connID uint64) bool {
+	mp.mu.Lock()
+	defer mp.mu.Unlock()
+	return mp.removedConnections[connID]
+}
+
+func (mp *mockPool) Len() int {
+	return 0
+}
+
+func (mp *mockPool) IdleLen() int {
+	return 0
+}
+
+func (mp *mockPool) Stats() *pool.Stats {
+	return &pool.Stats{}
+}
+
+func (mp *mockPool) AddPoolHook(hook pool.PoolHook) {
+	// Mock implementation - do nothing
+}
+
+func (mp *mockPool) RemovePoolHook(hook pool.PoolHook) {
+	// Mock implementation - do nothing
+}
+
+func (mp *mockPool) Close() error {
+	return nil
+}
+
+// TestConnectionHook tests the Redis connection processor functionality
+func TestConnectionHook(t *testing.T) {
+	// Create a base dialer for testing
+	baseDialer := func(ctx context.Context, network, addr string) (net.Conn, error) {
+		return &mockNetConn{addr: addr}, nil
+	}
+
+	t.Run("SuccessfulEventDrivenHandoff", func(t *testing.T) {
+		config := &Config{
+			Mode:              MaintNotificationsAuto,
+			EndpointType:      EndpointTypeAuto,
+			MaxWorkers:        1,  // Use only 1 worker to ensure synchronization
+			HandoffQueueSize:  10, // Explicit queue size to avoid 0-size queue
+			MaxHandoffRetries: 3,
+			LogLevel:          2,
+		}
+		processor := NewPoolHook(baseDialer, "tcp", config, nil)
+		defer processor.Shutdown(context.Background())
+
+		conn := createMockPoolConnection()
+		if err := conn.MarkForHandoff("new-endpoint:6379", 12345); err != nil {
+			t.Fatalf("Failed to mark connection for handoff: %v", err)
+		}
+
+		// Verify connection is marked for handoff
+		if !conn.ShouldHandoff() {
+			t.Fatal("Connection should be marked for handoff")
+		}
+		// Set a mock initialization function with synchronization
+		initConnCalled := make(chan bool, 1)
+		proceedWithInit := make(chan bool, 1)
+		initConnFunc := func(ctx context.Context, cn *pool.Conn) error {
+			select {
+			case initConnCalled <- true:
+			default:
+			}
+			// Wait for test to proceed
+			<-proceedWithInit
+			return nil
+		}
+		conn.SetInitConnFunc(initConnFunc)
+
+		ctx := context.Background()
+		shouldPool, shouldRemove, err := processor.OnPut(ctx, conn)
+		if err != nil {
+			t.Errorf("OnPut should not error: %v", err)
+		}
+
+		// Should pool the connection immediately (handoff queued)
+		if !shouldPool {
+			t.Error("Connection should be pooled immediately with event-driven handoff")
+		}
+		if shouldRemove {
+			t.Error("Connection should not be removed when queuing handoff")
+		}
+
+		// Wait for initialization to be called (indicates handoff started)
+		select {
+		case <-initConnCalled:
+			// Good, initialization was called
+		case <-time.After(1 * time.Second):
+			t.Fatal("Timeout waiting for initialization function to be called")
+		}
+
+		// Connection should be in pending map while initialization is blocked
+		if _, pending := processor.GetPendingMap().Load(conn.GetID()); !pending {
+			t.Error("Connection should be in pending handoffs map")
+		}
+
+		// Allow initialization to proceed
+		proceedWithInit <- true
+
+		// Wait for handoff to complete with proper timeout and polling
+		timeout := time.After(2 * time.Second)
+		ticker := time.NewTicker(10 * time.Millisecond)
+		defer ticker.Stop()
+
+		handoffCompleted := false
+		for !handoffCompleted {
+			select {
+			case <-timeout:
+				t.Fatal("Timeout waiting for handoff to complete")
+			case <-ticker.C:
+				if _, pending := processor.GetPendingMap().Load(conn); !pending {
+					handoffCompleted = true
+				}
+			}
+		}
+
+		// Verify handoff completed (removed from pending map)
+		if _, pending := processor.GetPendingMap().Load(conn); pending {
+			t.Error("Connection should be removed from pending map after handoff")
+		}
+
+		// Verify connection is usable again
+		if !conn.IsUsable() {
+			t.Error("Connection should be usable after successful handoff")
+		}
+
+		// Verify handoff state is cleared
+		if conn.ShouldHandoff() {
+			t.Error("Connection should not be marked for handoff after completion")
+		}
+	})
+
+	t.Run("HandoffNotNeeded", func(t *testing.T) {
+		processor := NewPoolHook(baseDialer, "tcp", nil, nil)
+		conn := createMockPoolConnection()
+		// Don't mark for handoff
+
+		ctx := context.Background()
+		shouldPool, shouldRemove, err := processor.OnPut(ctx, conn)
+		if err != nil {
+			t.Errorf("OnPut should not error when handoff not needed: %v", err)
+		}
+
+		// Should pool the connection normally
+		if !shouldPool {
+			t.Error("Connection should be pooled when no handoff needed")
+		}
+		if shouldRemove {
+			t.Error("Connection should not be removed when no handoff needed")
+		}
+	})
+
+	t.Run("EmptyEndpoint", func(t *testing.T) {
+		processor := NewPoolHook(baseDialer, "tcp", nil, nil)
+		conn := createMockPoolConnection()
+		if err := conn.MarkForHandoff("", 12345); err != nil { // Empty endpoint
+			t.Fatalf("Failed to mark connection for handoff: %v", err)
+		}
+
+		ctx := context.Background()
+		shouldPool, shouldRemove, err := processor.OnPut(ctx, conn)
+		if err != nil {
+			t.Errorf("OnPut should not error with empty endpoint: %v", err)
+		}
+
+		// Should pool the connection (empty endpoint clears state)
+		if !shouldPool {
+			t.Error("Connection should be pooled after clearing empty endpoint")
+		}
+		if shouldRemove {
+			t.Error("Connection should not be removed after clearing empty endpoint")
+		}
+
+		// State should be cleared
+		if conn.ShouldHandoff() {
+			t.Error("Connection should not be marked for handoff after clearing empty endpoint")
+		}
+	})
+
+	t.Run("EventDrivenHandoffDialerError", func(t *testing.T) {
+		// Create a failing base dialer
+		failingDialer := func(ctx context.Context, network, addr string) (net.Conn, error) {
+			return nil, errors.New("dial failed")
+		}
+
+		config := &Config{
+			Mode:              MaintNotificationsAuto,
+			EndpointType:      EndpointTypeAuto,
+			MaxWorkers:        2,
+			HandoffQueueSize:  10,
+			MaxHandoffRetries: 2,                      // Reduced retries for faster test
+			HandoffTimeout:    500 * time.Millisecond, // Shorter timeout for faster test
+			LogLevel:          2,
+		}
+		processor := NewPoolHook(failingDialer, "tcp", config, nil)
+		defer processor.Shutdown(context.Background())
+
+		conn := createMockPoolConnection()
+		if err := conn.MarkForHandoff("new-endpoint:6379", 12345); err != nil {
+			t.Fatalf("Failed to mark connection for handoff: %v", err)
+		}
+
+		ctx := context.Background()
+		shouldPool, shouldRemove, err := processor.OnPut(ctx, conn)
+		if err != nil {
+			t.Errorf("OnPut should not return error to caller: %v", err)
+		}
+
+		// Should pool the connection initially (handoff queued)
+		if !shouldPool {
+			t.Error("Connection should be pooled initially with event-driven handoff")
+		}
+		if shouldRemove {
+			t.Error("Connection should not be removed when queuing handoff")
+		}
+
+		// Wait for handoff to complete and fail with proper timeout and polling
+		timeout := time.After(3 * time.Second)
+		ticker := time.NewTicker(10 * time.Millisecond)
+		defer ticker.Stop()
+
+		// wait for handoff to start
+		time.Sleep(50 * time.Millisecond)
+		handoffCompleted := false
+		for !handoffCompleted {
+			select {
+			case <-timeout:
+				t.Fatal("Timeout waiting for failed handoff to complete")
+			case <-ticker.C:
+				if _, pending := processor.GetPendingMap().Load(conn.GetID()); !pending {
+					handoffCompleted = true
+				}
+			}
+		}
+
+		// Connection should be removed from pending map after failed handoff
+		if _, pending := processor.GetPendingMap().Load(conn.GetID()); pending {
+			t.Error("Connection should be removed from pending map after failed handoff")
+		}
+
+		// Wait for retries to complete (with MaxHandoffRetries=2, it will retry twice then give up)
+		// Each retry has a delay of handoffTimeout/2 = 250ms, so wait for all retries to complete
+		time.Sleep(800 * time.Millisecond)
+
+		// After max retries are reached, the connection should be removed from pool
+		// and handoff state should be cleared
+		if conn.ShouldHandoff() {
+			t.Error("Connection should not be marked for handoff after max retries reached")
+		}
+
+		t.Logf("EventDrivenHandoffDialerError test completed successfully")
+	})
+
+	t.Run("BufferedDataRESP2", func(t *testing.T) {
+		processor := NewPoolHook(baseDialer, "tcp", nil, nil)
+		conn := createMockPoolConnection()
+
+		// For this test, we'll just verify the logic works for connections without buffered data
+		// The actual buffered data detection is handled by the pool's connection health check
+		// which is outside the scope of the Redis connection processor
+
+		ctx := context.Background()
+		shouldPool, shouldRemove, err := processor.OnPut(ctx, conn)
+		if err != nil {
+			t.Errorf("OnPut should not error: %v", err)
+		}
+
+		// Should pool the connection normally (no buffered data in mock)
+		if !shouldPool {
+			t.Error("Connection should be pooled when no buffered data")
+		}
+		if shouldRemove {
+			t.Error("Connection should not be removed when no buffered data")
+		}
+	})
+
+	t.Run("OnGet", func(t *testing.T) {
+		processor := NewPoolHook(baseDialer, "tcp", nil, nil)
+		conn := createMockPoolConnection()
+
+		ctx := context.Background()
+		err := processor.OnGet(ctx, conn, false)
+		if err != nil {
+			t.Errorf("OnGet should not error for normal connection: %v", err)
+		}
+	})
+
+	t.Run("OnGetWithPendingHandoff", func(t *testing.T) {
+		config := &Config{
+			Mode:              MaintNotificationsAuto,
+			EndpointType:      EndpointTypeAuto,
+			MaxWorkers:        2,
+			HandoffQueueSize:  10,
+			MaxHandoffRetries: 3, // Explicit queue size to avoid 0-size queue
+			LogLevel:          2,
+		}
+		processor := NewPoolHook(baseDialer, "tcp", config, nil)
+		defer processor.Shutdown(context.Background())
+
+		conn := createMockPoolConnection()
+
+		// Simulate a pending handoff by marking for handoff and queuing
+		conn.MarkForHandoff("new-endpoint:6379", 12345)
+		processor.GetPendingMap().Store(conn.GetID(), int64(12345)) // Store connID -> seqID
+		conn.MarkQueuedForHandoff()                                 // Mark as queued (sets usable=false)
+
+		ctx := context.Background()
+		err := processor.OnGet(ctx, conn, false)
+		if err != ErrConnectionMarkedForHandoff {
+			t.Errorf("Expected ErrConnectionMarkedForHandoff, got %v", err)
+		}
+
+		// Clean up
+		processor.GetPendingMap().Delete(conn)
+	})
+
+	t.Run("EventDrivenStateManagement", func(t *testing.T) {
+		processor := NewPoolHook(baseDialer, "tcp", nil, nil)
+		defer processor.Shutdown(context.Background())
+
+		conn := createMockPoolConnection()
+
+		// Test initial state - no pending handoffs
+		if _, pending := processor.GetPendingMap().Load(conn); pending {
+			t.Error("New connection should not have pending handoffs")
+		}
+
+		// Test adding to pending map
+		conn.MarkForHandoff("new-endpoint:6379", 12345)
+		processor.GetPendingMap().Store(conn.GetID(), int64(12345)) // Store connID -> seqID
+		conn.MarkQueuedForHandoff()                                 // Mark as queued (sets usable=false)
+
+		if _, pending := processor.GetPendingMap().Load(conn.GetID()); !pending {
+			t.Error("Connection should be in pending map")
+		}
+
+		// Test OnGet with pending handoff
+		ctx := context.Background()
+		err := processor.OnGet(ctx, conn, false)
+		if err != ErrConnectionMarkedForHandoff {
+			t.Error("Should return ErrConnectionMarkedForHandoff for pending connection")
+		}
+
+		// Test removing from pending map and clearing handoff state
+		processor.GetPendingMap().Delete(conn)
+		if _, pending := processor.GetPendingMap().Load(conn); pending {
+			t.Error("Connection should be removed from pending map")
+		}
+
+		// Clear handoff state to simulate completed handoff
+		conn.ClearHandoffState()
+		conn.SetUsable(true) // Make connection usable again
+
+		// Test OnGet without pending handoff
+		err = processor.OnGet(ctx, conn, false)
+		if err != nil {
+			t.Errorf("Should not return error for non-pending connection: %v", err)
+		}
+	})
+
+	t.Run("EventDrivenQueueOptimization", func(t *testing.T) {
+		// Create processor with small queue to test optimization features
+		config := &Config{
+			MaxWorkers:        3,
+			HandoffQueueSize:  2,
+			MaxHandoffRetries: 3, // Small queue to trigger optimizations
+			LogLevel:          3, // Debug level to see optimization logs
+		}
+
+		baseDialer := func(ctx context.Context, network, addr string) (net.Conn, error) {
+			// Add small delay to simulate network latency
+			time.Sleep(10 * time.Millisecond)
+			return &mockNetConn{addr: addr}, nil
+		}
+
+		processor := NewPoolHook(baseDialer, "tcp", config, nil)
+		defer processor.Shutdown(context.Background())
+
+		// Create multiple connections that need handoff to fill the queue
+		connections := make([]*pool.Conn, 5)
+		for i := 0; i < 5; i++ {
+			connections[i] = createMockPoolConnection()
+			if err := connections[i].MarkForHandoff("new-endpoint:6379", int64(i)); err != nil {
+				t.Fatalf("Failed to mark conn[%d] for handoff: %v", i, err)
+			}
+			// Set a mock initialization function
+			connections[i].SetInitConnFunc(func(ctx context.Context, cn *pool.Conn) error {
+				return nil
+			})
+		}
+
+		ctx := context.Background()
+		successCount := 0
+
+		// Process connections - should trigger scaling and timeout logic
+		for _, conn := range connections {
+			shouldPool, shouldRemove, err := processor.OnPut(ctx, conn)
+			if err != nil {
+				t.Logf("OnPut returned error (expected with timeout): %v", err)
+			}
+
+			if shouldPool && !shouldRemove {
+				successCount++
+			}
+		}
+
+		// With timeout and scaling, most handoffs should eventually succeed
+		if successCount == 0 {
+			t.Error("Should have queued some handoffs with timeout and scaling")
+		}
+
+		t.Logf("Successfully queued %d handoffs with optimization features", successCount)
+
+		// Give time for workers to process and scaling to occur
+		time.Sleep(100 * time.Millisecond)
+	})
+
+	t.Run("WorkerScalingBehavior", func(t *testing.T) {
+		// Create processor with small queue to test scaling behavior
+		config := &Config{
+			MaxWorkers:        15, // Set to >= 10 to test explicit value preservation
+			HandoffQueueSize:  1,
+			MaxHandoffRetries: 3, // Very small queue to force scaling
+			LogLevel:          2, // Info level to see scaling logs
+		}
+
+		processor := NewPoolHook(baseDialer, "tcp", config, nil)
+		defer processor.Shutdown(context.Background())
+
+		// Verify initial worker count (should be 0 with on-demand workers)
+		if processor.GetCurrentWorkers() != 0 {
+			t.Errorf("Expected 0 initial workers with on-demand system, got %d", processor.GetCurrentWorkers())
+		}
+		if processor.GetMaxWorkers() != 15 {
+			t.Errorf("Expected maxWorkers=15, got %d", processor.GetMaxWorkers())
+		}
+
+		// The on-demand worker behavior creates workers only when needed
+		// This test just verifies the basic configuration is correct
+		t.Logf("On-demand worker configuration verified - Max: %d, Current: %d",
+			processor.GetMaxWorkers(), processor.GetCurrentWorkers())
+	})
+
+	t.Run("PassiveTimeoutRestoration", func(t *testing.T) {
+		// Create processor with fast post-handoff duration for testing
+		config := &Config{
+			MaxWorkers:                 2,
+			HandoffQueueSize:           10,
+			MaxHandoffRetries:          3,                      // Allow retries for successful handoff
+			PostHandoffRelaxedDuration: 100 * time.Millisecond, // Fast expiration for testing
+			RelaxedTimeout:             5 * time.Second,
+			LogLevel:                   2,
+		}
+
+		processor := NewPoolHook(baseDialer, "tcp", config, nil)
+		defer processor.Shutdown(context.Background())
+
+		ctx := context.Background()
+
+		// Create a connection and trigger handoff
+		conn := createMockPoolConnection()
+		if err := conn.MarkForHandoff("new-endpoint:6379", 1); err != nil {
+			t.Fatalf("Failed to mark connection for handoff: %v", err)
+		}
+
+		// Set a mock initialization function
+		conn.SetInitConnFunc(func(ctx context.Context, cn *pool.Conn) error {
+			return nil
+		})
+
+		// Process the connection to trigger handoff
+		shouldPool, shouldRemove, err := processor.OnPut(ctx, conn)
+		if err != nil {
+			t.Errorf("Handoff should succeed: %v", err)
+		}
+		if !shouldPool || shouldRemove {
+			t.Error("Connection should be pooled after handoff")
+		}
+
+		// Wait for handoff to complete with proper timeout and polling
+		timeout := time.After(1 * time.Second)
+		ticker := time.NewTicker(5 * time.Millisecond)
+		defer ticker.Stop()
+
+		handoffCompleted := false
+		for !handoffCompleted {
+			select {
+			case <-timeout:
+				t.Fatal("Timeout waiting for handoff to complete")
+			case <-ticker.C:
+				if _, pending := processor.GetPendingMap().Load(conn); !pending {
+					handoffCompleted = true
+				}
+			}
+		}
+
+		// Verify relaxed timeout is set with deadline
+		if !conn.HasRelaxedTimeout() {
+			t.Error("Connection should have relaxed timeout after handoff")
+		}
+
+		// Test that timeout is still active before deadline
+		// We'll use HasRelaxedTimeout which internally checks the deadline
+		if !conn.HasRelaxedTimeout() {
+			t.Error("Connection should still have active relaxed timeout before deadline")
+		}
+
+		// Wait for deadline to pass
+		time.Sleep(150 * time.Millisecond) // 100ms deadline + buffer
+
+		// Test that timeout is automatically restored after deadline
+		// HasRelaxedTimeout should return false after deadline passes
+		if conn.HasRelaxedTimeout() {
+			t.Error("Connection should not have active relaxed timeout after deadline")
+		}
+
+		// Additional verification: calling HasRelaxedTimeout again should still return false
+		// and should have cleared the internal timeout values
+		if conn.HasRelaxedTimeout() {
+			t.Error("Connection should not have relaxed timeout after deadline (second check)")
+		}
+
+		t.Logf("Passive timeout restoration test completed successfully")
+	})
+
+	t.Run("UsableFlagBehavior", func(t *testing.T) {
+		config := &Config{
+			MaxWorkers:        2,
+			HandoffQueueSize:  10,
+			MaxHandoffRetries: 3,
+			LogLevel:          2,
+		}
+
+		processor := NewPoolHook(baseDialer, "tcp", config, nil)
+		defer processor.Shutdown(context.Background())
+
+		ctx := context.Background()
+
+		// Create a new connection without setting it usable
+		mockNetConn := &mockNetConn{addr: "test:6379"}
+		conn := pool.NewConn(mockNetConn)
+
+		// Initially, connection should not be usable (not initialized)
+		if conn.IsUsable() {
+			t.Error("New connection should not be usable before initialization")
+		}
+
+		// Simulate initialization by setting usable to true
+		conn.SetUsable(true)
+		if !conn.IsUsable() {
+			t.Error("Connection should be usable after initialization")
+		}
+
+		// OnGet should succeed for usable connection
+		err := processor.OnGet(ctx, conn, false)
+		if err != nil {
+			t.Errorf("OnGet should succeed for usable connection: %v", err)
+		}
+
+		// Mark connection for handoff
+		if err := conn.MarkForHandoff("new-endpoint:6379", 1); err != nil {
+			t.Fatalf("Failed to mark connection for handoff: %v", err)
+		}
+
+		// Set a mock initialization function
+		conn.SetInitConnFunc(func(ctx context.Context, cn *pool.Conn) error {
+			return nil
+		})
+
+		// Connection should still be usable until queued, but marked for handoff
+		if !conn.IsUsable() {
+			t.Error("Connection should still be usable after being marked for handoff (until queued)")
+		}
+		if !conn.ShouldHandoff() {
+			t.Error("Connection should be marked for handoff")
+		}
+
+		// OnGet should fail for connection marked for handoff
+		err = processor.OnGet(ctx, conn, false)
+		if err == nil {
+			t.Error("OnGet should fail for connection marked for handoff")
+		}
+		if err != ErrConnectionMarkedForHandoff {
+			t.Errorf("Expected ErrConnectionMarkedForHandoff, got %v", err)
+		}
+
+		// Process the connection to trigger handoff
+		shouldPool, shouldRemove, err := processor.OnPut(ctx, conn)
+		if err != nil {
+			t.Errorf("OnPut should succeed: %v", err)
+		}
+		if !shouldPool || shouldRemove {
+			t.Error("Connection should be pooled after handoff")
+		}
+
+		// Wait for handoff to complete
+		time.Sleep(50 * time.Millisecond)
+
+		// After handoff completion, connection should be usable again
+		if !conn.IsUsable() {
+			t.Error("Connection should be usable after handoff completion")
+		}
+
+		// OnGet should succeed again
+		err = processor.OnGet(ctx, conn, false)
+		if err != nil {
+			t.Errorf("OnGet should succeed after handoff completion: %v", err)
+		}
+
+		t.Logf("Usable flag behavior test completed successfully")
+	})
+
+	t.Run("StaticQueueBehavior", func(t *testing.T) {
+		config := &Config{
+			MaxWorkers:        3,
+			HandoffQueueSize:  50,
+			MaxHandoffRetries: 3, // Explicit static queue size
+			LogLevel:          2,
+		}
+
+		processor := NewPoolHookWithPoolSize(baseDialer, "tcp", config, nil, 100) // Pool size: 100
+		defer processor.Shutdown(context.Background())
+
+		// Verify queue capacity matches configured size
+		queueCapacity := cap(processor.GetHandoffQueue())
+		if queueCapacity != 50 {
+			t.Errorf("Expected queue capacity 50, got %d", queueCapacity)
+		}
+
+		// Test that queue size is static regardless of pool size
+		// (No dynamic resizing should occur)
+
+		ctx := context.Background()
+
+		// Fill part of the queue
+		for i := 0; i < 10; i++ {
+			conn := createMockPoolConnection()
+			if err := conn.MarkForHandoff("new-endpoint:6379", int64(i+1)); err != nil {
+				t.Fatalf("Failed to mark conn[%d] for handoff: %v", i, err)
+			}
+			// Set a mock initialization function
+			conn.SetInitConnFunc(func(ctx context.Context, cn *pool.Conn) error {
+				return nil
+			})
+
+			shouldPool, shouldRemove, err := processor.OnPut(ctx, conn)
+			if err != nil {
+				t.Errorf("Failed to queue handoff %d: %v", i, err)
+			}
+
+			if !shouldPool || shouldRemove {
+				t.Errorf("conn[%d] should be pooled after handoff (shouldPool=%v, shouldRemove=%v)",
+					i, shouldPool, shouldRemove)
+			}
+		}
+
+		// Verify queue capacity remains static (the main purpose of this test)
+		finalCapacity := cap(processor.GetHandoffQueue())
+
+		if finalCapacity != 50 {
+			t.Errorf("Queue capacity should remain static at 50, got %d", finalCapacity)
+		}
+
+		// Note: We don't check queue size here because workers process items quickly
+		// The important thing is that the capacity remains static regardless of pool size
+	})
+
+	t.Run("ConnectionRemovalOnHandoffFailure", func(t *testing.T) {
+		// Create a failing dialer that will cause handoff initialization to fail
+		failingDialer := func(ctx context.Context, network, addr string) (net.Conn, error) {
+			// Return a connection that will fail during initialization
+			return &mockNetConn{addr: addr, shouldFailInit: true}, nil
+		}
+
+		config := &Config{
+			MaxWorkers:        2,
+			HandoffQueueSize:  10,
+			MaxHandoffRetries: 3,
+			LogLevel:          2,
+		}
+
+		processor := NewPoolHook(failingDialer, "tcp", config, nil)
+		defer processor.Shutdown(context.Background())
+
+		// Create a mock pool that tracks removals
+		mockPool := &mockPool{removedConnections: make(map[uint64]bool)}
+		processor.SetPool(mockPool)
+
+		ctx := context.Background()
+
+		// Create a connection and mark it for handoff
+		conn := createMockPoolConnection()
+		if err := conn.MarkForHandoff("new-endpoint:6379", 1); err != nil {
+			t.Fatalf("Failed to mark connection for handoff: %v", err)
+		}
+
+		// Set a failing initialization function
+		conn.SetInitConnFunc(func(ctx context.Context, cn *pool.Conn) error {
+			return fmt.Errorf("initialization failed")
+		})
+
+		// Process the connection - handoff should fail and connection should be removed
+		shouldPool, shouldRemove, err := processor.OnPut(ctx, conn)
+		if err != nil {
+			t.Errorf("OnPut should not error: %v", err)
+		}
+		if !shouldPool || shouldRemove {
+			t.Error("Connection should be pooled after failed handoff attempt")
+		}
+
+		// Wait for handoff to be attempted and fail
+		time.Sleep(100 * time.Millisecond)
+
+		// Verify that the connection was removed from the pool
+		if !mockPool.WasRemoved(conn.GetID()) {
+			t.Errorf("conn[%d] should have been removed from pool after handoff failure", conn.GetID())
+		}
+
+		t.Logf("Connection removal on handoff failure test completed successfully")
+	})
+
+	t.Run("PostHandoffRelaxedTimeout", func(t *testing.T) {
+		// Create config with short post-handoff duration for testing
+		config := &Config{
+			MaxWorkers:                 2,
+			HandoffQueueSize:           10,
+			MaxHandoffRetries:          3, // Allow retries for successful handoff
+			RelaxedTimeout:             5 * time.Second,
+			PostHandoffRelaxedDuration: 100 * time.Millisecond, // Short for testing
+		}
+
+		baseDialer := func(ctx context.Context, network, addr string) (net.Conn, error) {
+			return &mockNetConn{addr: addr}, nil
+		}
+
+		processor := NewPoolHook(baseDialer, "tcp", config, nil)
+		defer processor.Shutdown(context.Background())
+
+		conn := createMockPoolConnection()
+		if err := conn.MarkForHandoff("new-endpoint:6379", 12345); err != nil {
+			t.Fatalf("Failed to mark connection for handoff: %v", err)
+		}
+
+		// Set a mock initialization function
+		conn.SetInitConnFunc(func(ctx context.Context, cn *pool.Conn) error {
+			return nil
+		})
+
+		ctx := context.Background()
+		shouldPool, shouldRemove, err := processor.OnPut(ctx, conn)
+
+		if err != nil {
+			t.Fatalf("OnPut failed: %v", err)
+		}
+
+		if !shouldPool {
+			t.Error("Connection should be pooled after successful handoff")
+		}
+
+		if shouldRemove {
+			t.Error("Connection should not be removed after successful handoff")
+		}
+
+		// Wait for the handoff to complete (it happens asynchronously)
+		timeout := time.After(1 * time.Second)
+		ticker := time.NewTicker(5 * time.Millisecond)
+		defer ticker.Stop()
+
+		handoffCompleted := false
+		for !handoffCompleted {
+			select {
+			case <-timeout:
+				t.Fatal("Timeout waiting for handoff to complete")
+			case <-ticker.C:
+				if _, pending := processor.GetPendingMap().Load(conn); !pending {
+					handoffCompleted = true
+				}
+			}
+		}
+
+		// Verify that relaxed timeout was applied to the new connection
+		if !conn.HasRelaxedTimeout() {
+			t.Error("New connection should have relaxed timeout applied after handoff")
+		}
+
+		// Wait for the post-handoff duration to expire
+		time.Sleep(150 * time.Millisecond) // Slightly longer than PostHandoffRelaxedDuration
+
+		// Verify that relaxed timeout was automatically cleared
+		if conn.HasRelaxedTimeout() {
+			t.Error("Relaxed timeout should be automatically cleared after post-handoff duration")
+		}
+	})
+
+	t.Run("MarkForHandoff returns error when already marked", func(t *testing.T) {
+		conn := createMockPoolConnection()
+
+		// First mark should succeed
+		if err := conn.MarkForHandoff("new-endpoint:6379", 1); err != nil {
+			t.Fatalf("First MarkForHandoff should succeed: %v", err)
+		}
+
+		// Second mark should fail
+		if err := conn.MarkForHandoff("another-endpoint:6379", 2); err == nil {
+			t.Fatal("Second MarkForHandoff should return error")
+		} else if err.Error() != "connection is already marked for handoff" {
+			t.Fatalf("Expected specific error message, got: %v", err)
+		}
+
+		// Verify original handoff data is preserved
+		if !conn.ShouldHandoff() {
+			t.Fatal("Connection should still be marked for handoff")
+		}
+		if conn.GetHandoffEndpoint() != "new-endpoint:6379" {
+			t.Fatalf("Expected original endpoint, got: %s", conn.GetHandoffEndpoint())
+		}
+		if conn.GetMovingSeqID() != 1 {
+			t.Fatalf("Expected original sequence ID, got: %d", conn.GetMovingSeqID())
+		}
+	})
+
+	t.Run("HandoffTimeoutConfiguration", func(t *testing.T) {
+		// Test that HandoffTimeout from config is actually used
+		customTimeout := 2 * time.Second
+		config := &Config{
+			MaxWorkers:        2,
+			HandoffQueueSize:  10,
+			HandoffTimeout:    customTimeout, // Custom timeout
+			MaxHandoffRetries: 1,             // Single retry to speed up test
+			LogLevel:          2,
+		}
+
+		processor := NewPoolHook(baseDialer, "tcp", config, nil)
+		defer processor.Shutdown(context.Background())
+
+		// Create a connection that will test the timeout
+		conn := createMockPoolConnection()
+		if err := conn.MarkForHandoff("test-endpoint:6379", 123); err != nil {
+			t.Fatalf("Failed to mark connection for handoff: %v", err)
+		}
+
+		// Set a dialer that will check the context timeout
+		var timeoutVerified int32 // Use atomic for thread safety
+		conn.SetInitConnFunc(func(ctx context.Context, cn *pool.Conn) error {
+			// Check that the context has the expected timeout
+			deadline, ok := ctx.Deadline()
+			if !ok {
+				t.Error("Context should have a deadline")
+				return errors.New("no deadline")
+			}
+
+			// The deadline should be approximately customTimeout from now
+			expectedDeadline := time.Now().Add(customTimeout)
+			timeDiff := deadline.Sub(expectedDeadline)
+			if timeDiff < -500*time.Millisecond || timeDiff > 500*time.Millisecond {
+				t.Errorf("Context deadline not as expected. Expected around %v, got %v (diff: %v)",
+					expectedDeadline, deadline, timeDiff)
+			} else {
+				atomic.StoreInt32(&timeoutVerified, 1)
+			}
+
+			return nil // Successful handoff
+		})
+
+		// Trigger handoff
+		shouldPool, shouldRemove, err := processor.OnPut(context.Background(), conn)
+		if err != nil {
+			t.Errorf("OnPut should not return error: %v", err)
+		}
+
+		// Connection should be queued for handoff
+		if !shouldPool || shouldRemove {
+			t.Errorf("Connection should be pooled for handoff processing")
+		}
+
+		// Wait for handoff to complete
+		time.Sleep(500 * time.Millisecond)
+
+		if atomic.LoadInt32(&timeoutVerified) == 0 {
+			t.Error("HandoffTimeout was not properly applied to context")
+		}
+
+		t.Logf("HandoffTimeout configuration test completed successfully")
+	})
+}

hitless/push_notification_handler.go

@@ -0,0 +1,276 @@
+package hitless
+
+import (
+	"context"
+	"fmt"
+	"time"
+
+	"github.com/redis/go-redis/v9/internal"
+	"github.com/redis/go-redis/v9/internal/pool"
+	"github.com/redis/go-redis/v9/push"
+)
+
+// NotificationHandler handles push notifications for the simplified manager.
+type NotificationHandler struct {
+	manager *HitlessManager
+}
+
+// HandlePushNotification processes push notifications with hook support.
+func (snh *NotificationHandler) HandlePushNotification(ctx context.Context, handlerCtx push.NotificationHandlerContext, notification []interface{}) error {
+	if len(notification) == 0 {
+		internal.Logger.Printf(ctx, "hitless: invalid notification format: %v", notification)
+		return ErrInvalidNotification
+	}
+
+	notificationType, ok := notification[0].(string)
+	if !ok {
+		internal.Logger.Printf(ctx, "hitless: invalid notification type format: %v", notification[0])
+		return ErrInvalidNotification
+	}
+
+	// Process pre-hooks - they can modify the notification or skip processing
+	modifiedNotification, shouldContinue := snh.manager.processPreHooks(ctx, handlerCtx, notificationType, notification)
+	if !shouldContinue {
+		return nil // Hooks decided to skip processing
+	}
+
+	var err error
+	switch notificationType {
+	case NotificationMoving:
+		err = snh.handleMoving(ctx, handlerCtx, modifiedNotification)
+	case NotificationMigrating:
+		err = snh.handleMigrating(ctx, handlerCtx, modifiedNotification)
+	case NotificationMigrated:
+		err = snh.handleMigrated(ctx, handlerCtx, modifiedNotification)
+	case NotificationFailingOver:
+		err = snh.handleFailingOver(ctx, handlerCtx, modifiedNotification)
+	case NotificationFailedOver:
+		err = snh.handleFailedOver(ctx, handlerCtx, modifiedNotification)
+	default:
+		// Ignore other notification types (e.g., pub/sub messages)
+		err = nil
+	}
+
+	// Process post-hooks with the result
+	snh.manager.processPostHooks(ctx, handlerCtx, notificationType, modifiedNotification, err)
+
+	return err
+}
+
+// handleMoving processes MOVING notifications.
+// ["MOVING", seqNum, timeS, endpoint] - per-connection handoff
+func (snh *NotificationHandler) handleMoving(ctx context.Context, handlerCtx push.NotificationHandlerContext, notification []interface{}) error {
+	if len(notification) < 3 {
+		internal.Logger.Printf(ctx, "hitless: invalid MOVING notification: %v", notification)
+		return ErrInvalidNotification
+	}
+	seqID, ok := notification[1].(int64)
+	if !ok {
+		internal.Logger.Printf(ctx, "hitless: invalid seqID in MOVING notification: %v", notification[1])
+		return ErrInvalidNotification
+	}
+
+	// Extract timeS
+	timeS, ok := notification[2].(int64)
+	if !ok {
+		internal.Logger.Printf(ctx, "hitless: invalid timeS in MOVING notification: %v", notification[2])
+		return ErrInvalidNotification
+	}
+
+	newEndpoint := ""
+	if len(notification) > 3 {
+		// Extract new endpoint
+		newEndpoint, ok = notification[3].(string)
+		if !ok {
+			internal.Logger.Printf(ctx, "hitless: invalid newEndpoint in MOVING notification: %v", notification[3])
+			return ErrInvalidNotification
+		}
+	}
+
+	// Get the connection that received this notification
+	conn := handlerCtx.Conn
+	if conn == nil {
+		internal.Logger.Printf(ctx, "hitless: no connection in handler context for MOVING notification")
+		return ErrInvalidNotification
+	}
+
+	// Type assert to get the underlying pool connection
+	var poolConn *pool.Conn
+	if pc, ok := conn.(*pool.Conn); ok {
+		poolConn = pc
+	} else {
+		internal.Logger.Printf(ctx, "hitless: invalid connection type in handler context for MOVING notification - %T %#v", conn, handlerCtx)
+		return ErrInvalidNotification
+	}
+
+	// If the connection is closed or not pooled, we can ignore the notification
+	// this connection won't be remembered by the pool and will be garbage collected
+	// Keep pubsub connections around since they are not pooled but are long-lived
+	// and should be allowed to handoff (the pubsub instance will reconnect and change
+	// the underlying *pool.Conn)
+	if (poolConn.IsClosed() || !poolConn.IsPooled()) && !poolConn.IsPubSub() {
+		return nil
+	}
+
+	deadline := time.Now().Add(time.Duration(timeS) * time.Second)
+	// If newEndpoint is empty, we should schedule a handoff to the current endpoint in timeS/2 seconds
+	if newEndpoint == "" || newEndpoint == internal.RedisNull {
+		if snh.manager.config.LogLevel.DebugOrAbove() { // Debug level
+			internal.Logger.Printf(ctx, "hitless: conn[%d] scheduling handoff to current endpoint in %v seconds",
+				poolConn.GetID(), timeS/2)
+		}
+		// same as current endpoint
+		newEndpoint = snh.manager.options.GetAddr()
+		// delay the handoff for timeS/2 seconds to the same endpoint
+		// do this in a goroutine to avoid blocking the notification handler
+		// NOTE: This timer is started while parsing the notification, so the connection is not marked for handoff
+		// and there should be no possibility of a race condition or double handoff.
+		time.AfterFunc(time.Duration(timeS/2)*time.Second, func() {
+			if poolConn == nil || poolConn.IsClosed() {
+				return
+			}
+			if err := snh.markConnForHandoff(poolConn, newEndpoint, seqID, deadline); err != nil {
+				// Log error but don't fail the goroutine - use background context since original may be cancelled
+				internal.Logger.Printf(context.Background(), "hitless: failed to mark connection for handoff: %v", err)
+			}
+		})
+		return nil
+	}
+
+	return snh.markConnForHandoff(poolConn, newEndpoint, seqID, deadline)
+}
+
+func (snh *NotificationHandler) markConnForHandoff(conn *pool.Conn, newEndpoint string, seqID int64, deadline time.Time) error {
+	if err := conn.MarkForHandoff(newEndpoint, seqID); err != nil {
+		internal.Logger.Printf(context.Background(), "hitless: failed to mark connection for handoff: %v", err)
+		// Connection is already marked for handoff, which is acceptable
+		// This can happen if multiple MOVING notifications are received for the same connection
+		return nil
+	}
+	// Optionally track in hitless manager for monitoring/debugging
+	if snh.manager != nil {
+		connID := conn.GetID()
+		// Track the operation (ignore errors since this is optional)
+		_ = snh.manager.TrackMovingOperationWithConnID(context.Background(), newEndpoint, deadline, seqID, connID)
+	} else {
+		return fmt.Errorf("hitless: manager not initialized")
+	}
+	return nil
+}
+
+// handleMigrating processes MIGRATING notifications.
+func (snh *NotificationHandler) handleMigrating(ctx context.Context, handlerCtx push.NotificationHandlerContext, notification []interface{}) error {
+	// MIGRATING notifications indicate that a connection is about to be migrated
+	// Apply relaxed timeouts to the specific connection that received this notification
+	if len(notification) < 2 {
+		internal.Logger.Printf(ctx, "hitless: invalid MIGRATING notification: %v", notification)
+		return ErrInvalidNotification
+	}
+
+	if handlerCtx.Conn == nil {
+		internal.Logger.Printf(ctx, "hitless: no connection in handler context for MIGRATING notification")
+		return ErrInvalidNotification
+	}
+
+	conn, ok := handlerCtx.Conn.(*pool.Conn)
+	if !ok {
+		internal.Logger.Printf(ctx, "hitless: invalid connection type in handler context for MIGRATING notification")
+		return ErrInvalidNotification
+	}
+
+	// Apply relaxed timeout to this specific connection
+	if snh.manager.config.LogLevel.InfoOrAbove() { // Debug level
+		internal.Logger.Printf(ctx, "hitless: conn[%d] applying relaxed timeout (%v) for MIGRATING notification",
+			conn.GetID(),
+			snh.manager.config.RelaxedTimeout)
+	}
+	conn.SetRelaxedTimeout(snh.manager.config.RelaxedTimeout, snh.manager.config.RelaxedTimeout)
+	return nil
+}
+
+// handleMigrated processes MIGRATED notifications.
+func (snh *NotificationHandler) handleMigrated(ctx context.Context, handlerCtx push.NotificationHandlerContext, notification []interface{}) error {
+	// MIGRATED notifications indicate that a connection migration has completed
+	// Restore normal timeouts for the specific connection that received this notification
+	if len(notification) < 2 {
+		internal.Logger.Printf(ctx, "hitless: invalid MIGRATED notification: %v", notification)
+		return ErrInvalidNotification
+	}
+
+	if handlerCtx.Conn == nil {
+		internal.Logger.Printf(ctx, "hitless: no connection in handler context for MIGRATED notification")
+		return ErrInvalidNotification
+	}
+
+	conn, ok := handlerCtx.Conn.(*pool.Conn)
+	if !ok {
+		internal.Logger.Printf(ctx, "hitless: invalid connection type in handler context for MIGRATED notification")
+		return ErrInvalidNotification
+	}
+
+	// Clear relaxed timeout for this specific connection
+	if snh.manager.config.LogLevel.InfoOrAbove() { // Debug level
+		connID := conn.GetID()
+		internal.Logger.Printf(ctx, "hitless: conn[%d] clearing relaxed timeout for MIGRATED notification", connID)
+	}
+	conn.ClearRelaxedTimeout()
+	return nil
+}
+
+// handleFailingOver processes FAILING_OVER notifications.
+func (snh *NotificationHandler) handleFailingOver(ctx context.Context, handlerCtx push.NotificationHandlerContext, notification []interface{}) error {
+	// FAILING_OVER notifications indicate that a connection is about to failover
+	// Apply relaxed timeouts to the specific connection that received this notification
+	if len(notification) < 2 {
+		internal.Logger.Printf(ctx, "hitless: invalid FAILING_OVER notification: %v", notification)
+		return ErrInvalidNotification
+	}
+
+	if handlerCtx.Conn == nil {
+		internal.Logger.Printf(ctx, "hitless: no connection in handler context for FAILING_OVER notification")
+		return ErrInvalidNotification
+	}
+
+	conn, ok := handlerCtx.Conn.(*pool.Conn)
+	if !ok {
+		internal.Logger.Printf(ctx, "hitless: invalid connection type in handler context for FAILING_OVER notification")
+		return ErrInvalidNotification
+	}
+
+	// Apply relaxed timeout to this specific connection
+	if snh.manager.config.LogLevel.InfoOrAbove() { // Debug level
+		connID := conn.GetID()
+		internal.Logger.Printf(ctx, "hitless: conn[%d] applying relaxed timeout (%v) for FAILING_OVER notification", connID, snh.manager.config.RelaxedTimeout)
+	}
+	conn.SetRelaxedTimeout(snh.manager.config.RelaxedTimeout, snh.manager.config.RelaxedTimeout)
+	return nil
+}
+
+// handleFailedOver processes FAILED_OVER notifications.
+func (snh *NotificationHandler) handleFailedOver(ctx context.Context, handlerCtx push.NotificationHandlerContext, notification []interface{}) error {
+	// FAILED_OVER notifications indicate that a connection failover has completed
+	// Restore normal timeouts for the specific connection that received this notification
+	if len(notification) < 2 {
+		internal.Logger.Printf(ctx, "hitless: invalid FAILED_OVER notification: %v", notification)
+		return ErrInvalidNotification
+	}
+
+	if handlerCtx.Conn == nil {
+		internal.Logger.Printf(ctx, "hitless: no connection in handler context for FAILED_OVER notification")
+		return ErrInvalidNotification
+	}
+
+	conn, ok := handlerCtx.Conn.(*pool.Conn)
+	if !ok {
+		internal.Logger.Printf(ctx, "hitless: invalid connection type in handler context for FAILED_OVER notification")
+		return ErrInvalidNotification
+	}
+
+	// Clear relaxed timeout for this specific connection
+	if snh.manager.config.LogLevel.InfoOrAbove() { // Debug level
+		connID := conn.GetID()
+		internal.Logger.Printf(ctx, "hitless: conn[%d] clearing relaxed timeout for FAILED_OVER notification", connID)
+	}
+	conn.ClearRelaxedTimeout()
+	return nil
+}

hitless/state.go

@@ -0,0 +1,24 @@
+package hitless
+
+// State represents the current state of a hitless upgrade operation.
+type State int
+
+const (
+	// StateIdle indicates no upgrade is in progress
+	StateIdle State = iota
+
+	// StateHandoff indicates a connection handoff is in progress
+	StateMoving
+)
+
+// String returns a string representation of the state.
+func (s State) String() string {
+	switch s {
+	case StateIdle:
+		return "idle"
+	case StateMoving:
+		return "moving"
+	default:
+		return "unknown"
+	}
+}

hset_benchmark_test.go

@@ -0,0 +1,245 @@
+package redis_test
+
+import (
+	"context"
+	"fmt"
+	"testing"
+	"time"
+
+	"github.com/redis/go-redis/v9"
+)
+
+// HSET Benchmark Tests
+//
+// This file contains benchmark tests for Redis HSET operations with different scales:
+// 1, 10, 100, 1000, 10000, 100000 operations
+//
+// Prerequisites:
+// - Redis server running on localhost:6379
+// - No authentication required
+//
+// Usage:
+//   go test -bench=BenchmarkHSET -v ./hset_benchmark_test.go
+//   go test -bench=BenchmarkHSETPipelined -v ./hset_benchmark_test.go
+//   go test -bench=. -v ./hset_benchmark_test.go  # Run all benchmarks
+//
+// Example output:
+//   BenchmarkHSET/HSET_1_operations-8         	    5000	    250000 ns/op	1000000.00 ops/sec
+//   BenchmarkHSET/HSET_100_operations-8       	     100	  10000000 ns/op	 100000.00 ops/sec
+//
+// The benchmarks test three different approaches:
+// 1. Individual HSET commands (BenchmarkHSET)
+// 2. Pipelined HSET commands (BenchmarkHSETPipelined)
+
+// BenchmarkHSET benchmarks HSET operations with different scales
+func BenchmarkHSET(b *testing.B) {
+	ctx := context.Background()
+
+	// Setup Redis client
+	rdb := redis.NewClient(&redis.Options{
+		Addr: "localhost:6379",
+		DB:   0,
+	})
+	defer rdb.Close()
+
+	// Test connection
+	if err := rdb.Ping(ctx).Err(); err != nil {
+		b.Skipf("Redis server not available: %v", err)
+	}
+
+	// Clean up before and after tests
+	defer func() {
+		rdb.FlushDB(ctx)
+	}()
+
+	scales := []int{1, 10, 100, 1000, 10000, 100000}
+
+	for _, scale := range scales {
+		b.Run(fmt.Sprintf("HSET_%d_operations", scale), func(b *testing.B) {
+			benchmarkHSETOperations(b, rdb, ctx, scale)
+		})
+	}
+}
+
+// benchmarkHSETOperations performs the actual HSET benchmark for a given scale
+func benchmarkHSETOperations(b *testing.B, rdb *redis.Client, ctx context.Context, operations int) {
+	hashKey := fmt.Sprintf("benchmark_hash_%d", operations)
+
+	b.ResetTimer()
+	b.StartTimer()
+	totalTimes := []time.Duration{}
+
+	for i := 0; i < b.N; i++ {
+		b.StopTimer()
+		// Clean up the hash before each iteration
+		rdb.Del(ctx, hashKey)
+		b.StartTimer()
+
+		startTime := time.Now()
+		// Perform the specified number of HSET operations
+		for j := 0; j < operations; j++ {
+			field := fmt.Sprintf("field_%d", j)
+			value := fmt.Sprintf("value_%d", j)
+
+			err := rdb.HSet(ctx, hashKey, field, value).Err()
+			if err != nil {
+				b.Fatalf("HSET operation failed: %v", err)
+			}
+		}
+		totalTimes = append(totalTimes, time.Now().Sub(startTime))
+	}
+
+	// Stop the timer to calculate metrics
+	b.StopTimer()
+
+	// Report operations per second
+	opsPerSec := float64(operations*b.N) / b.Elapsed().Seconds()
+	b.ReportMetric(opsPerSec, "ops/sec")
+
+	// Report average time per operation
+	avgTimePerOp := b.Elapsed().Nanoseconds() / int64(operations*b.N)
+	b.ReportMetric(float64(avgTimePerOp), "ns/op")
+	// report average time in milliseconds from totalTimes
+	avgTimePerOpMs := totalTimes[0].Milliseconds() / int64(len(totalTimes))
+	b.ReportMetric(float64(avgTimePerOpMs), "ms")
+}
+
+// BenchmarkHSETPipelined benchmarks HSET operations using pipelining for better performance
+func BenchmarkHSETPipelined(b *testing.B) {
+	ctx := context.Background()
+
+	// Setup Redis client
+	rdb := redis.NewClient(&redis.Options{
+		Addr: "localhost:6379",
+		DB:   0,
+	})
+	defer rdb.Close()
+
+	// Test connection
+	if err := rdb.Ping(ctx).Err(); err != nil {
+		b.Skipf("Redis server not available: %v", err)
+	}
+
+	// Clean up before and after tests
+	defer func() {
+		rdb.FlushDB(ctx)
+	}()
+
+	scales := []int{1, 10, 100, 1000, 10000, 100000}
+
+	for _, scale := range scales {
+		b.Run(fmt.Sprintf("HSET_Pipelined_%d_operations", scale), func(b *testing.B) {
+			benchmarkHSETPipelined(b, rdb, ctx, scale)
+		})
+	}
+}
+
+// benchmarkHSETPipelined performs HSET benchmark using pipelining
+func benchmarkHSETPipelined(b *testing.B, rdb *redis.Client, ctx context.Context, operations int) {
+	hashKey := fmt.Sprintf("benchmark_hash_pipelined_%d", operations)
+
+	b.ResetTimer()
+	b.StartTimer()
+	totalTimes := []time.Duration{}
+
+	for i := 0; i < b.N; i++ {
+		b.StopTimer()
+		// Clean up the hash before each iteration
+		rdb.Del(ctx, hashKey)
+		b.StartTimer()
+
+		startTime := time.Now()
+		// Use pipelining for better performance
+		pipe := rdb.Pipeline()
+
+		// Add all HSET operations to the pipeline
+		for j := 0; j < operations; j++ {
+			field := fmt.Sprintf("field_%d", j)
+			value := fmt.Sprintf("value_%d", j)
+			pipe.HSet(ctx, hashKey, field, value)
+		}
+
+		// Execute all operations at once
+		_, err := pipe.Exec(ctx)
+		if err != nil {
+			b.Fatalf("Pipeline execution failed: %v", err)
+		}
+		totalTimes = append(totalTimes, time.Now().Sub(startTime))
+	}
+
+	b.StopTimer()
+
+	// Report operations per second
+	opsPerSec := float64(operations*b.N) / b.Elapsed().Seconds()
+	b.ReportMetric(opsPerSec, "ops/sec")
+
+	// Report average time per operation
+	avgTimePerOp := b.Elapsed().Nanoseconds() / int64(operations*b.N)
+	b.ReportMetric(float64(avgTimePerOp), "ns/op")
+	// report average time in milliseconds from totalTimes
+	avgTimePerOpMs := totalTimes[0].Milliseconds() / int64(len(totalTimes))
+	b.ReportMetric(float64(avgTimePerOpMs), "ms")
+}
+
+// add same tests but with RESP2
+func BenchmarkHSET_RESP2(b *testing.B) {
+	ctx := context.Background()
+
+	// Setup Redis client
+	rdb := redis.NewClient(&redis.Options{
+		Addr:     "localhost:6379",
+		Password: "", // no password docs
+		DB:       0,  // use default DB
+		Protocol: 2,
+	})
+	defer rdb.Close()
+
+	// Test connection
+	if err := rdb.Ping(ctx).Err(); err != nil {
+		b.Skipf("Redis server not available: %v", err)
+	}
+
+	// Clean up before and after tests
+	defer func() {
+		rdb.FlushDB(ctx)
+	}()
+
+	scales := []int{1, 10, 100, 1000, 10000, 100000}
+
+	for _, scale := range scales {
+		b.Run(fmt.Sprintf("HSET_RESP2_%d_operations", scale), func(b *testing.B) {
+			benchmarkHSETOperations(b, rdb, ctx, scale)
+		})
+	}
+}
+
+func BenchmarkHSETPipelined_RESP2(b *testing.B) {
+	ctx := context.Background()
+
+	// Setup Redis client
+	rdb := redis.NewClient(&redis.Options{
+		Addr:     "localhost:6379",
+		Password: "", // no password docs
+		DB:       0,  // use default DB
+		Protocol: 2,
+	})
+	defer rdb.Close()
+
+	// Test connection
+	if err := rdb.Ping(ctx).Err(); err != nil {
+		b.Skipf("Redis server not available: %v", err)
+	}
+
+	// Clean up before and after tests
+	defer func() {
+		rdb.FlushDB(ctx)
+	}()
+
+	scales := []int{1, 10, 100, 1000, 10000, 100000}
+
+	for _, scale := range scales {
+		b.Run(fmt.Sprintf("HSET_Pipelined_RESP2_%d_operations", scale), func(b *testing.B) {
+			benchmarkHSETPipelined(b, rdb, ctx, scale)
+		})
+	}
+}

internal/interfaces/interfaces.go

@@ -0,0 +1,54 @@
+// Package interfaces provides shared interfaces used by both the main redis package
+// and the hitless upgrade package to avoid circular dependencies.
+package interfaces
+
+import (
+	"context"
+	"net"
+	"time"
+)
+
+// NotificationProcessor is (most probably) a push.NotificationProcessor
+// forward declaration to avoid circular imports
+type NotificationProcessor interface {
+	RegisterHandler(pushNotificationName string, handler interface{}, protected bool) error
+	UnregisterHandler(pushNotificationName string) error
+	GetHandler(pushNotificationName string) interface{}
+}
+
+// ClientInterface defines the interface that clients must implement for hitless upgrades.
+type ClientInterface interface {
+	// GetOptions returns the client options.
+	GetOptions() OptionsInterface
+
+	// GetPushProcessor returns the client's push notification processor.
+	GetPushProcessor() NotificationProcessor
+}
+
+// OptionsInterface defines the interface for client options.
+// Uses an adapter pattern to avoid circular dependencies.
+type OptionsInterface interface {
+	// GetReadTimeout returns the read timeout.
+	GetReadTimeout() time.Duration
+
+	// GetWriteTimeout returns the write timeout.
+	GetWriteTimeout() time.Duration
+
+	// GetNetwork returns the network type.
+	GetNetwork() string
+
+	// GetAddr returns the connection address.
+	GetAddr() string
+
+	// IsTLSEnabled returns true if TLS is enabled.
+	IsTLSEnabled() bool
+
+	// GetProtocol returns the protocol version.
+	GetProtocol() int
+
+	// GetPoolSize returns the connection pool size.
+	GetPoolSize() int
+
+	// NewDialer returns a new dialer function for the connection.
+	NewDialer() func(context.Context) (net.Conn, error)
+}

internal/log.go

@@ -7,20 +7,27 @@ import (
 	"os"
 )
 
+// TODO (ned): Revisit logging
+// Add more standardized approach with log levels and configurability
+
 type Logging interface {
 	Printf(ctx context.Context, format string, v ...interface{})
 }
 
-type logger struct {
+type DefaultLogger struct {
 	log *log.Logger
 }
 
-func (l *logger) Printf(ctx context.Context, format string, v ...interface{}) {
+func (l *DefaultLogger) Printf(ctx context.Context, format string, v ...interface{}) {
 	_ = l.log.Output(2, fmt.Sprintf(format, v...))
 }
 
+func NewDefaultLogger() Logging {
+	return &DefaultLogger{
+		log: log.New(os.Stderr, "redis: ", log.LstdFlags|log.Lshortfile),
+	}
+}
+
 // Logger calls Output to print to the stderr.
 // Arguments are handled in the manner of fmt.Print.
-var Logger Logging = &logger{
-	log: log.New(os.Stderr, "redis: ", log.LstdFlags|log.Lshortfile),
-}
+var Logger Logging = NewDefaultLogger()

internal/pool/bench_test.go

@@ -2,6 +2,7 @@ package pool_test
 
 import (
 	"context"
+	"errors"
 	"fmt"
 	"testing"
 	"time"
@@ -31,7 +32,7 @@ func BenchmarkPoolGetPut(b *testing.B) {
 		b.Run(bm.String(), func(b *testing.B) {
 			connPool := pool.NewConnPool(&pool.Options{
 				Dialer:          dummyDialer,
-				PoolSize:        bm.poolSize,
+				PoolSize:        int32(bm.poolSize),
 				PoolTimeout:     time.Second,
 				DialTimeout:     1 * time.Second,
 				ConnMaxIdleTime: time.Hour,
@@ -75,7 +76,7 @@ func BenchmarkPoolGetRemove(b *testing.B) {
 		b.Run(bm.String(), func(b *testing.B) {
 			connPool := pool.NewConnPool(&pool.Options{
 				Dialer:          dummyDialer,
-				PoolSize:        bm.poolSize,
+				PoolSize:        int32(bm.poolSize),
 				PoolTimeout:     time.Second,
 				DialTimeout:     1 * time.Second,
 				ConnMaxIdleTime: time.Hour,
@@ -89,7 +90,7 @@ func BenchmarkPoolGetRemove(b *testing.B) {
 					if err != nil {
 						b.Fatal(err)
 					}
-					connPool.Remove(ctx, cn, nil)
+					connPool.Remove(ctx, cn, errors.New("Bench test remove"))
 				}
 			})
 		})

internal/pool/buffer_size_test.go

@@ -26,7 +26,7 @@ var _ = Describe("Buffer Size Configuration", func() {
 	It("should use default buffer sizes when not specified", func() {
 		connPool = pool.NewConnPool(&pool.Options{
 			Dialer:      dummyDialer,
-			PoolSize:    1,
+			PoolSize:    int32(1),
 			PoolTimeout: 1000,
 		})
 
@@ -48,7 +48,7 @@ var _ = Describe("Buffer Size Configuration", func() {
 
 		connPool = pool.NewConnPool(&pool.Options{
 			Dialer:          dummyDialer,
-			PoolSize:        1,
+			PoolSize:        int32(1),
 			PoolTimeout:     1000,
 			ReadBufferSize:  customReadSize,
 			WriteBufferSize: customWriteSize,
@@ -69,7 +69,7 @@ var _ = Describe("Buffer Size Configuration", func() {
 	It("should handle zero buffer sizes by using defaults", func() {
 		connPool = pool.NewConnPool(&pool.Options{
 			Dialer:          dummyDialer,
-			PoolSize:        1,
+			PoolSize:        int32(1),
 			PoolTimeout:     1000,
 			ReadBufferSize:  0, // Should use default
 			WriteBufferSize: 0, // Should use default
@@ -105,7 +105,7 @@ var _ = Describe("Buffer Size Configuration", func() {
 		// without setting ReadBufferSize and WriteBufferSize
 		connPool = pool.NewConnPool(&pool.Options{
 			Dialer:      dummyDialer,
-			PoolSize:    1,
+			PoolSize:    int32(1),
 			PoolTimeout: 1000,
 			// ReadBufferSize and WriteBufferSize are not set (will be 0)
 		})

internal/pool/conn.go

@@ -3,7 +3,10 @@ package pool
 import (
 	"bufio"
 	"context"
+	"errors"
+	"fmt"
 	"net"
+	"sync"
 	"sync/atomic"
 	"time"
 
@@ -12,17 +15,74 @@ import (
 
 var noDeadline = time.Time{}
 
+// Global atomic counter for connection IDs
+var connIDCounter uint64
+
+// HandoffState represents the atomic state for connection handoffs
+// This struct is stored atomically to prevent race conditions between
+// checking handoff status and reading handoff parameters
+type HandoffState struct {
+	ShouldHandoff bool   // Whether connection should be handed off
+	Endpoint      string // New endpoint for handoff
+	SeqID         int64  // Sequence ID from MOVING notification
+}
+
+// atomicNetConn is a wrapper to ensure consistent typing in atomic.Value
+type atomicNetConn struct {
+	conn net.Conn
+}
+
+// generateConnID generates a fast unique identifier for a connection with zero allocations
+func generateConnID() uint64 {
+	return atomic.AddUint64(&connIDCounter, 1)
+}
+
 type Conn struct {
 	usedAt int64 // atomic
-	netConn net.Conn
+
+	// Lock-free netConn access using atomic.Value
+	// Contains *atomicNetConn wrapper, accessed atomically for better performance
+	netConnAtomic atomic.Value // stores *atomicNetConn
 
 	rd *proto.Reader
 	bw *bufio.Writer
 	wr *proto.Writer
 
-	Inited    bool
+	// Lightweight mutex to protect reader operations during handoff
+	// Only used for the brief period during SetNetConn and HasBufferedData/PeekReplyTypeSafe
+	readerMu sync.RWMutex
+
+	Inited    atomic.Bool
 	pooled    bool
+	pubsub    bool
+	closed    atomic.Bool
 	createdAt time.Time
+	expiresAt time.Time
+
+	// Hitless upgrade support: relaxed timeouts during migrations/failovers
+	// Using atomic operations for lock-free access to avoid mutex contention
+	relaxedReadTimeoutNs  atomic.Int64 // time.Duration as nanoseconds
+	relaxedWriteTimeoutNs atomic.Int64 // time.Duration as nanoseconds
+	relaxedDeadlineNs     atomic.Int64 // time.Time as nanoseconds since epoch
+
+	// Counter to track multiple relaxed timeout setters if we have nested calls
+	// will be decremented when ClearRelaxedTimeout is called or deadline is reached
+	// if counter reaches 0, we clear the relaxed timeouts
+	relaxedCounter atomic.Int32
+
+	// Connection initialization function for reconnections
+	initConnFunc func(context.Context, *Conn) error
+
+	// Connection identifier for unique tracking across handoffs
+	id uint64 // Unique numeric identifier for this connection
+
+	// Handoff state - using atomic operations for lock-free access
+	usableAtomic         atomic.Bool   // Connection usability state
+	handoffRetriesAtomic atomic.Uint32 // Retry counter for handoff attempts
+
+	// Atomic handoff state to prevent race conditions
+	// Stores *HandoffState to ensure atomic updates of all handoff-related fields
+	handoffStateAtomic atomic.Value // stores *HandoffState
 
 	onClose func() error
 }
@@ -33,8 +93,8 @@ func NewConn(netConn net.Conn) *Conn {
 
 func NewConnWithBufferSize(netConn net.Conn, readBufSize, writeBufSize int) *Conn {
 	cn := &Conn{
-		netConn:   netConn,
 		createdAt: time.Now(),
+		id:        generateConnID(), // Generate unique ID for this connection
 	}
 
 	// Use specified buffer sizes, or fall back to 32KiB defaults if 0
@@ -50,6 +110,21 @@ func NewConnWithBufferSize(netConn net.Conn, readBufSize, writeBufSize int) *Con
 		cn.bw = bufio.NewWriterSize(netConn, proto.DefaultBufferSize)
 	}
 
+	// Store netConn atomically for lock-free access using wrapper
+	cn.netConnAtomic.Store(&atomicNetConn{conn: netConn})
+
+	// Initialize atomic state
+	cn.usableAtomic.Store(false)        // false initially, set to true after initialization
+	cn.handoffRetriesAtomic.Store(0)    // 0 initially
+
+	// Initialize handoff state atomically
+	initialHandoffState := &HandoffState{
+		ShouldHandoff: false,
+		Endpoint:      "",
+		SeqID:         0,
+	}
+	cn.handoffStateAtomic.Store(initialHandoffState)
+
 	cn.wr = proto.NewWriter(cn.bw)
 	cn.SetUsedAt(time.Now())
 	return cn
@@ -64,23 +139,430 @@ func (cn *Conn) SetUsedAt(tm time.Time) {
 	atomic.StoreInt64(&cn.usedAt, tm.Unix())
 }
 
+// getNetConn returns the current network connection using atomic load (lock-free).
+// This is the fast path for accessing netConn without mutex overhead.
+func (cn *Conn) getNetConn() net.Conn {
+	if v := cn.netConnAtomic.Load(); v != nil {
+		if wrapper, ok := v.(*atomicNetConn); ok {
+			return wrapper.conn
+		}
+	}
+	return nil
+}
+
+// setNetConn stores the network connection atomically (lock-free).
+// This is used for the fast path of connection replacement.
+func (cn *Conn) setNetConn(netConn net.Conn) {
+	cn.netConnAtomic.Store(&atomicNetConn{conn: netConn})
+}
+
+// Lock-free helper methods for handoff state management
+
+// isUsable returns true if the connection is safe to use (lock-free).
+func (cn *Conn) isUsable() bool {
+	return cn.usableAtomic.Load()
+}
+
+// setUsable sets the usable flag atomically (lock-free).
+func (cn *Conn) setUsable(usable bool) {
+	cn.usableAtomic.Store(usable)
+}
+
+// getHandoffState returns the current handoff state atomically (lock-free).
+func (cn *Conn) getHandoffState() *HandoffState {
+	state := cn.handoffStateAtomic.Load()
+	if state == nil {
+		// Return default state if not initialized
+		return &HandoffState{
+			ShouldHandoff: false,
+			Endpoint:      "",
+			SeqID:         0,
+		}
+	}
+	return state.(*HandoffState)
+}
+
+// setHandoffState sets the handoff state atomically (lock-free).
+func (cn *Conn) setHandoffState(state *HandoffState) {
+	cn.handoffStateAtomic.Store(state)
+}
+
+// shouldHandoff returns true if connection needs handoff (lock-free).
+func (cn *Conn) shouldHandoff() bool {
+	return cn.getHandoffState().ShouldHandoff
+}
+
+// getMovingSeqID returns the sequence ID atomically (lock-free).
+func (cn *Conn) getMovingSeqID() int64 {
+	return cn.getHandoffState().SeqID
+}
+
+// getNewEndpoint returns the new endpoint atomically (lock-free).
+func (cn *Conn) getNewEndpoint() string {
+	return cn.getHandoffState().Endpoint
+}
+
+// setHandoffRetries sets the retry count atomically (lock-free).
+func (cn *Conn) setHandoffRetries(retries int) {
+	cn.handoffRetriesAtomic.Store(uint32(retries))
+}
+
+// incrementHandoffRetries atomically increments and returns the new retry count (lock-free).
+func (cn *Conn) incrementHandoffRetries(delta int) int {
+	return int(cn.handoffRetriesAtomic.Add(uint32(delta)))
+}
+
+// IsUsable returns true if the connection is safe to use for new commands (lock-free).
+func (cn *Conn) IsUsable() bool {
+	return cn.isUsable()
+}
+
+// IsPooled returns true if the connection is managed by a pool and will be pooled on Put.
+func (cn *Conn) IsPooled() bool {
+	return cn.pooled
+}
+
+// IsPubSub returns true if the connection is used for PubSub.
+func (cn *Conn) IsPubSub() bool {
+	return cn.pubsub
+}
+
+func (cn *Conn) IsInited() bool {
+	return cn.Inited.Load()
+}
+
+// SetUsable sets the usable flag for the connection (lock-free).
+func (cn *Conn) SetUsable(usable bool) {
+	cn.setUsable(usable)
+}
+
+// SetRelaxedTimeout sets relaxed timeouts for this connection during hitless upgrades.
+// These timeouts will be used for all subsequent commands until the deadline expires.
+// Uses atomic operations for lock-free access.
+func (cn *Conn) SetRelaxedTimeout(readTimeout, writeTimeout time.Duration) {
+	cn.relaxedCounter.Add(1)
+	cn.relaxedReadTimeoutNs.Store(int64(readTimeout))
+	cn.relaxedWriteTimeoutNs.Store(int64(writeTimeout))
+}
+
+// SetRelaxedTimeoutWithDeadline sets relaxed timeouts with an expiration deadline.
+// After the deadline, timeouts automatically revert to normal values.
+// Uses atomic operations for lock-free access.
+func (cn *Conn) SetRelaxedTimeoutWithDeadline(readTimeout, writeTimeout time.Duration, deadline time.Time) {
+	cn.SetRelaxedTimeout(readTimeout, writeTimeout)
+	cn.relaxedDeadlineNs.Store(deadline.UnixNano())
+}
+
+// ClearRelaxedTimeout removes relaxed timeouts, returning to normal timeout behavior.
+// Uses atomic operations for lock-free access.
+func (cn *Conn) ClearRelaxedTimeout() {
+	// Atomically decrement counter and check if we should clear
+	newCount := cn.relaxedCounter.Add(-1)
+	if newCount <= 0 {
+		// Use atomic load to get current value for CAS to avoid stale value race
+		current := cn.relaxedCounter.Load()
+		if current <= 0 && cn.relaxedCounter.CompareAndSwap(current, 0) {
+			cn.clearRelaxedTimeout()
+		}
+	}
+}
+
+func (cn *Conn) clearRelaxedTimeout() {
+	cn.relaxedReadTimeoutNs.Store(0)
+	cn.relaxedWriteTimeoutNs.Store(0)
+	cn.relaxedDeadlineNs.Store(0)
+	cn.relaxedCounter.Store(0)
+}
+
+// HasRelaxedTimeout returns true if relaxed timeouts are currently active on this connection.
+// This checks both the timeout values and the deadline (if set).
+// Uses atomic operations for lock-free access.
+func (cn *Conn) HasRelaxedTimeout() bool {
+	// Fast path: no relaxed timeouts are set
+	if cn.relaxedCounter.Load() <= 0 {
+		return false
+	}
+
+	readTimeoutNs := cn.relaxedReadTimeoutNs.Load()
+	writeTimeoutNs := cn.relaxedWriteTimeoutNs.Load()
+
+	// If no relaxed timeouts are set, return false
+	if readTimeoutNs <= 0 && writeTimeoutNs <= 0 {
+		return false
+	}
+
+	deadlineNs := cn.relaxedDeadlineNs.Load()
+	// If no deadline is set, relaxed timeouts are active
+	if deadlineNs == 0 {
+		return true
+	}
+
+	// If deadline is set, check if it's still in the future
+	return time.Now().UnixNano() < deadlineNs
+}
+
+// getEffectiveReadTimeout returns the timeout to use for read operations.
+// If relaxed timeout is set and not expired, it takes precedence over the provided timeout.
+// This method automatically clears expired relaxed timeouts using atomic operations.
+func (cn *Conn) getEffectiveReadTimeout(normalTimeout time.Duration) time.Duration {
+	readTimeoutNs := cn.relaxedReadTimeoutNs.Load()
+
+	// Fast path: no relaxed timeout set
+	if readTimeoutNs <= 0 {
+		return normalTimeout
+	}
+
+	deadlineNs := cn.relaxedDeadlineNs.Load()
+	// If no deadline is set, use relaxed timeout
+	if deadlineNs == 0 {
+		return time.Duration(readTimeoutNs)
+	}
+
+	nowNs := time.Now().UnixNano()
+	// Check if deadline has passed
+	if nowNs < deadlineNs {
+		// Deadline is in the future, use relaxed timeout
+		return time.Duration(readTimeoutNs)
+	} else {
+		// Deadline has passed, clear relaxed timeouts atomically and use normal timeout
+		cn.relaxedCounter.Add(-1)
+		if cn.relaxedCounter.Load() <= 0 {
+			cn.clearRelaxedTimeout()
+		}
+		return normalTimeout
+	}
+}
+
+// getEffectiveWriteTimeout returns the timeout to use for write operations.
+// If relaxed timeout is set and not expired, it takes precedence over the provided timeout.
+// This method automatically clears expired relaxed timeouts using atomic operations.
+func (cn *Conn) getEffectiveWriteTimeout(normalTimeout time.Duration) time.Duration {
+	writeTimeoutNs := cn.relaxedWriteTimeoutNs.Load()
+
+	// Fast path: no relaxed timeout set
+	if writeTimeoutNs <= 0 {
+		return normalTimeout
+	}
+
+	deadlineNs := cn.relaxedDeadlineNs.Load()
+	// If no deadline is set, use relaxed timeout
+	if deadlineNs == 0 {
+		return time.Duration(writeTimeoutNs)
+	}
+
+	nowNs := time.Now().UnixNano()
+	// Check if deadline has passed
+	if nowNs < deadlineNs {
+		// Deadline is in the future, use relaxed timeout
+		return time.Duration(writeTimeoutNs)
+	} else {
+		// Deadline has passed, clear relaxed timeouts atomically and use normal timeout
+		cn.relaxedCounter.Add(-1)
+		if cn.relaxedCounter.Load() <= 0 {
+			cn.clearRelaxedTimeout()
+		}
+		return normalTimeout
+	}
+}
+
 func (cn *Conn) SetOnClose(fn func() error) {
 	cn.onClose = fn
 }
 
+// SetInitConnFunc sets the connection initialization function to be called on reconnections.
+func (cn *Conn) SetInitConnFunc(fn func(context.Context, *Conn) error) {
+	cn.initConnFunc = fn
+}
+
+// ExecuteInitConn runs the stored connection initialization function if available.
+func (cn *Conn) ExecuteInitConn(ctx context.Context) error {
+	if cn.initConnFunc != nil {
+		return cn.initConnFunc(ctx, cn)
+	}
+	return fmt.Errorf("redis: no initConnFunc set for conn[%d]", cn.GetID())
+}
+
 func (cn *Conn) SetNetConn(netConn net.Conn) {
-	cn.netConn = netConn
+	// Store the new connection atomically first (lock-free)
+	cn.setNetConn(netConn)
+	// Protect reader reset operations to avoid data races
+	// Use write lock since we're modifying the reader state
+	cn.readerMu.Lock()
 	cn.rd.Reset(netConn)
+	cn.readerMu.Unlock()
+
 	cn.bw.Reset(netConn)
 }
 
+// GetNetConn safely returns the current network connection using atomic load (lock-free).
+// This method is used by the pool for health checks and provides better performance.
+func (cn *Conn) GetNetConn() net.Conn {
+	return cn.getNetConn()
+}
+
+// SetNetConnAndInitConn replaces the underlying connection and executes the initialization.
+func (cn *Conn) SetNetConnAndInitConn(ctx context.Context, netConn net.Conn) error {
+	// New connection is not initialized yet
+	cn.Inited.Store(false)
+	// Replace the underlying connection
+	cn.SetNetConn(netConn)
+	return cn.ExecuteInitConn(ctx)
+}
+
+// MarkForHandoff marks the connection for handoff due to MOVING notification (lock-free).
+// Returns an error if the connection is already marked for handoff.
+// This method uses atomic compare-and-swap to ensure all handoff state is updated atomically.
+func (cn *Conn) MarkForHandoff(newEndpoint string, seqID int64) error {
+	const maxRetries = 50
+	const baseDelay = time.Microsecond
+
+	for attempt := 0; attempt < maxRetries; attempt++ {
+		currentState := cn.getHandoffState()
+
+		// Check if already marked for handoff
+		if currentState.ShouldHandoff {
+			return errors.New("connection is already marked for handoff")
+		}
+
+		// Create new state with handoff enabled
+		newState := &HandoffState{
+			ShouldHandoff: true,
+			Endpoint:      newEndpoint,
+			SeqID:         seqID,
+		}
+
+		// Atomic compare-and-swap to update entire state
+		if cn.handoffStateAtomic.CompareAndSwap(currentState, newState) {
+			return nil
+		}
+
+		// If CAS failed, add exponential backoff to reduce contention
+		if attempt < maxRetries-1 {
+			delay := baseDelay * time.Duration(1<<uint(attempt%10)) // Cap exponential growth
+			time.Sleep(delay)
+		}
+	}
+
+	return fmt.Errorf("failed to mark connection for handoff after %d attempts due to high contention", maxRetries)
+}
+
+func (cn *Conn) MarkQueuedForHandoff() error {
+	const maxRetries = 50
+	const baseDelay = time.Microsecond
+
+	for attempt := 0; attempt < maxRetries; attempt++ {
+		currentState := cn.getHandoffState()
+
+		// Check if marked for handoff
+		if !currentState.ShouldHandoff {
+			return errors.New("connection was not marked for handoff")
+		}
+
+		// Create new state with handoff disabled (queued)
+		newState := &HandoffState{
+			ShouldHandoff: false,
+			Endpoint:      currentState.Endpoint, // Preserve endpoint for handoff processing
+			SeqID:         currentState.SeqID,    // Preserve seqID for handoff processing
+		}
+
+		// Atomic compare-and-swap to update state
+		if cn.handoffStateAtomic.CompareAndSwap(currentState, newState) {
+			cn.setUsable(false)
+			return nil
+		}
+
+		// If CAS failed, add exponential backoff to reduce contention
+		if attempt < maxRetries-1 {
+			delay := baseDelay * time.Duration(1<<uint(attempt%10)) // Cap exponential growth
+			time.Sleep(delay)
+		}
+	}
+
+	return fmt.Errorf("failed to mark connection as queued for handoff after %d attempts due to high contention", maxRetries)
+}
+
+// ShouldHandoff returns true if the connection needs to be handed off (lock-free).
+func (cn *Conn) ShouldHandoff() bool {
+	return cn.shouldHandoff()
+}
+
+// GetHandoffEndpoint returns the new endpoint for handoff (lock-free).
+func (cn *Conn) GetHandoffEndpoint() string {
+	return cn.getNewEndpoint()
+}
+
+// GetMovingSeqID returns the sequence ID from the MOVING notification (lock-free).
+func (cn *Conn) GetMovingSeqID() int64 {
+	return cn.getMovingSeqID()
+}
+
+// GetHandoffInfo returns all handoff information atomically (lock-free).
+// This method prevents race conditions by returning all handoff state in a single atomic operation.
+// Returns (shouldHandoff, endpoint, seqID).
+func (cn *Conn) GetHandoffInfo() (bool, string, int64) {
+	state := cn.getHandoffState()
+	return state.ShouldHandoff, state.Endpoint, state.SeqID
+}
+
+// GetID returns the unique identifier for this connection.
+func (cn *Conn) GetID() uint64 {
+	return cn.id
+}
+
+// ClearHandoffState clears the handoff state after successful handoff (lock-free).
+func (cn *Conn) ClearHandoffState() {
+	// Create clean state
+	cleanState := &HandoffState{
+		ShouldHandoff: false,
+		Endpoint:      "",
+		SeqID:         0,
+	}
+
+	// Atomically set clean state
+	cn.setHandoffState(cleanState)
+	cn.setHandoffRetries(0)
+	cn.setUsable(true) // Connection is safe to use again after handoff completes
+}
+
+// IncrementAndGetHandoffRetries atomically increments and returns handoff retries (lock-free).
+func (cn *Conn) IncrementAndGetHandoffRetries(n int) int {
+	return cn.incrementHandoffRetries(n)
+}
+
+// HasBufferedData safely checks if the connection has buffered data.
+// This method is used to avoid data races when checking for push notifications.
+func (cn *Conn) HasBufferedData() bool {
+	// Use read lock for concurrent access to reader state
+	cn.readerMu.RLock()
+	defer cn.readerMu.RUnlock()
+	return cn.rd.Buffered() > 0
+}
+
+// PeekReplyTypeSafe safely peeks at the reply type.
+// This method is used to avoid data races when checking for push notifications.
+func (cn *Conn) PeekReplyTypeSafe() (byte, error) {
+	// Use read lock for concurrent access to reader state
+	cn.readerMu.RLock()
+	defer cn.readerMu.RUnlock()
+
+	if cn.rd.Buffered() <= 0 {
+		return 0, fmt.Errorf("redis: can't peek reply type, no data available")
+	}
+	return cn.rd.PeekReplyType()
+}
+
 func (cn *Conn) Write(b []byte) (int, error) {
-	return cn.netConn.Write(b)
+	// Lock-free netConn access for better performance
+	if netConn := cn.getNetConn(); netConn != nil {
+		return netConn.Write(b)
+	}
+	return 0, net.ErrClosed
 }
 
 func (cn *Conn) RemoteAddr() net.Addr {
-	if cn.netConn != nil {
-		return cn.netConn.RemoteAddr()
+	// Lock-free netConn access for better performance
+	if netConn := cn.getNetConn(); netConn != nil {
+		return netConn.RemoteAddr()
 	}
 	return nil
 }
@@ -89,7 +571,16 @@ func (cn *Conn) WithReader(
 	ctx context.Context, timeout time.Duration, fn func(rd *proto.Reader) error,
 ) error {
 	if timeout >= 0 {
-		if err := cn.netConn.SetReadDeadline(cn.deadline(ctx, timeout)); err != nil {
+		// Use relaxed timeout if set, otherwise use provided timeout
+		effectiveTimeout := cn.getEffectiveReadTimeout(timeout)
+
+		// Get the connection directly from atomic storage
+		netConn := cn.getNetConn()
+		if netConn == nil {
+			return fmt.Errorf("redis: connection not available")
+		}
+
+		if err := netConn.SetReadDeadline(cn.deadline(ctx, effectiveTimeout)); err != nil {
 			return err
 		}
 	}
@@ -100,13 +591,26 @@ func (cn *Conn) WithWriter(
 	ctx context.Context, timeout time.Duration, fn func(wr *proto.Writer) error,
 ) error {
 	if timeout >= 0 {
-		if err := cn.netConn.SetWriteDeadline(cn.deadline(ctx, timeout)); err != nil {
+		// Use relaxed timeout if set, otherwise use provided timeout
+		effectiveTimeout := cn.getEffectiveWriteTimeout(timeout)
+
+		// Always set write deadline, even if getNetConn() returns nil
+		// This prevents write operations from hanging indefinitely
+		if netConn := cn.getNetConn(); netConn != nil {
+			if err := netConn.SetWriteDeadline(cn.deadline(ctx, effectiveTimeout)); err != nil {
 				return err
 			}
+		} else {
+			// If getNetConn() returns nil, we still need to respect the timeout
+			// Return an error to prevent indefinite blocking
+			return fmt.Errorf("redis: connection not available for write operation")
+		}
 	}
 
 	if cn.bw.Buffered() > 0 {
-		cn.bw.Reset(cn.netConn)
+		if netConn := cn.getNetConn(); netConn != nil {
+			cn.bw.Reset(netConn)
+		}
 	}
 
 	if err := fn(cn.wr); err != nil {
@@ -116,12 +620,33 @@ func (cn *Conn) WithWriter(
 	return cn.bw.Flush()
 }
 
+func (cn *Conn) IsClosed() bool {
+	return cn.closed.Load()
+}
+
 func (cn *Conn) Close() error {
+	cn.closed.Store(true)
 	if cn.onClose != nil {
 		// ignore error
 		_ = cn.onClose()
 	}
-	return cn.netConn.Close()
+
+	// Lock-free netConn access for better performance
+	if netConn := cn.getNetConn(); netConn != nil {
+		return netConn.Close()
+	}
+	return nil
+}
+
+// MaybeHasData tries to peek at the next byte in the socket without consuming it
+// This is used to check if there are push notifications available
+// Important: This will work on Linux, but not on Windows
+func (cn *Conn) MaybeHasData() bool {
+	// Lock-free netConn access for better performance
+	if netConn := cn.getNetConn(); netConn != nil {
+		return maybeHasData(netConn)
+	}
+	return false
 }
 
 func (cn *Conn) deadline(ctx context.Context, timeout time.Duration) time.Time {

internal/pool/conn_check.go

@@ -12,6 +12,9 @@ import (
 
 var errUnexpectedRead = errors.New("unexpected read from socket")
 
+// connCheck checks if the connection is still alive and if there is data in the socket
+// it will try to peek at the next byte without consuming it since we may want to work with it
+// later on (e.g. push notifications)
 func connCheck(conn net.Conn) error {
 	// Reset previous timeout.
 	_ = conn.SetDeadline(time.Time{})
@@ -29,7 +32,9 @@ func connCheck(conn net.Conn) error {
 
 	if err := rawConn.Read(func(fd uintptr) bool {
 		var buf [1]byte
-		n, err := syscall.Read(int(fd), buf[:])
+		// Use MSG_PEEK to peek at data without consuming it
+		n, _, err := syscall.Recvfrom(int(fd), buf[:], syscall.MSG_PEEK|syscall.MSG_DONTWAIT)
+
 		switch {
 		case n == 0 && err == nil:
 			sysErr = io.EOF
@@ -47,3 +52,8 @@ func connCheck(conn net.Conn) error {
 
 	return sysErr
 }
+
+// maybeHasData checks if there is data in the socket without consuming it
+func maybeHasData(conn net.Conn) bool {
+	return connCheck(conn) == errUnexpectedRead
+}

internal/pool/conn_check_dummy.go

@@ -7,3 +7,8 @@ import "net"
 func connCheck(conn net.Conn) error {
 	return nil
 }
+
+// since we can't check for data on the socket, we just assume there is some
+func maybeHasData(conn net.Conn) bool {
+	return true
+}

internal/pool/conn_relaxed_timeout_test.go

@@ -0,0 +1,92 @@
+package pool
+
+import (
+	"net"
+	"sync"
+	"testing"
+	"time"
+)
+
+// TestConcurrentRelaxedTimeoutClearing tests the race condition fix in ClearRelaxedTimeout
+func TestConcurrentRelaxedTimeoutClearing(t *testing.T) {
+	// Create a dummy connection for testing
+	netConn := &net.TCPConn{}
+	cn := NewConn(netConn)
+	defer cn.Close()
+
+	// Set relaxed timeout multiple times to increase counter
+	cn.SetRelaxedTimeout(time.Second, time.Second)
+	cn.SetRelaxedTimeout(time.Second, time.Second)
+	cn.SetRelaxedTimeout(time.Second, time.Second)
+
+	// Verify counter is 3
+	if count := cn.relaxedCounter.Load(); count != 3 {
+		t.Errorf("Expected relaxed counter to be 3, got %d", count)
+	}
+
+	// Clear timeouts concurrently to test race condition fix
+	var wg sync.WaitGroup
+	for i := 0; i < 10; i++ {
+		wg.Add(1)
+		go func() {
+			defer wg.Done()
+			cn.ClearRelaxedTimeout()
+		}()
+	}
+	wg.Wait()
+
+	// Verify counter is 0 and timeouts are cleared
+	if count := cn.relaxedCounter.Load(); count != 0 {
+		t.Errorf("Expected relaxed counter to be 0 after clearing, got %d", count)
+	}
+	if timeout := cn.relaxedReadTimeoutNs.Load(); timeout != 0 {
+		t.Errorf("Expected relaxed read timeout to be 0, got %d", timeout)
+	}
+	if timeout := cn.relaxedWriteTimeoutNs.Load(); timeout != 0 {
+		t.Errorf("Expected relaxed write timeout to be 0, got %d", timeout)
+	}
+}
+
+// TestRelaxedTimeoutCounterRaceCondition tests the specific race condition scenario
+func TestRelaxedTimeoutCounterRaceCondition(t *testing.T) {
+	netConn := &net.TCPConn{}
+	cn := NewConn(netConn)
+	defer cn.Close()
+
+	// Set relaxed timeout once
+	cn.SetRelaxedTimeout(time.Second, time.Second)
+
+	// Verify counter is 1
+	if count := cn.relaxedCounter.Load(); count != 1 {
+		t.Errorf("Expected relaxed counter to be 1, got %d", count)
+	}
+
+	// Test concurrent clearing with race condition scenario
+	var wg sync.WaitGroup
+
+	// Multiple goroutines try to clear simultaneously
+	for i := 0; i < 5; i++ {
+		wg.Add(1)
+		go func() {
+			defer wg.Done()
+			cn.ClearRelaxedTimeout()
+		}()
+	}
+	wg.Wait()
+
+	// Verify final state is consistent
+	if count := cn.relaxedCounter.Load(); count != 0 {
+		t.Errorf("Expected relaxed counter to be 0 after concurrent clearing, got %d", count)
+	}
+	
+	// Verify timeouts are actually cleared
+	if timeout := cn.relaxedReadTimeoutNs.Load(); timeout != 0 {
+		t.Errorf("Expected relaxed read timeout to be cleared, got %d", timeout)
+	}
+	if timeout := cn.relaxedWriteTimeoutNs.Load(); timeout != 0 {
+		t.Errorf("Expected relaxed write timeout to be cleared, got %d", timeout)
+	}
+	if deadline := cn.relaxedDeadlineNs.Load(); deadline != 0 {
+		t.Errorf("Expected relaxed deadline to be cleared, got %d", deadline)
+	}
+}

internal/pool/export_test.go

@@ -10,7 +10,7 @@ func (cn *Conn) SetCreatedAt(tm time.Time) {
 }
 
 func (cn *Conn) NetConn() net.Conn {
-	return cn.netConn
+	return cn.getNetConn()
 }
 
 func (p *ConnPool) CheckMinIdleConns() {

internal/pool/hooks.go

@@ -0,0 +1,114 @@
+package pool
+
+import (
+	"context"
+	"sync"
+)
+
+// PoolHook defines the interface for connection lifecycle hooks.
+type PoolHook interface {
+	// OnGet is called when a connection is retrieved from the pool.
+	// It can modify the connection or return an error to prevent its use.
+	// It has isNewConn flag to indicate if this is a new connection (rather than idle from the pool)
+	// The flag can be used for gathering metrics on pool hit/miss ratio.
+	OnGet(ctx context.Context, conn *Conn, isNewConn bool) error
+
+	// OnPut is called when a connection is returned to the pool.
+	// It returns whether the connection should be pooled and whether it should be removed.
+	OnPut(ctx context.Context, conn *Conn) (shouldPool bool, shouldRemove bool, err error)
+}
+
+// PoolHookManager manages multiple pool hooks.
+type PoolHookManager struct {
+	hooks   []PoolHook
+	hooksMu sync.RWMutex
+}
+
+// NewPoolHookManager creates a new pool hook manager.
+func NewPoolHookManager() *PoolHookManager {
+	return &PoolHookManager{
+		hooks: make([]PoolHook, 0),
+	}
+}
+
+// AddHook adds a pool hook to the manager.
+// Hooks are called in the order they were added.
+func (phm *PoolHookManager) AddHook(hook PoolHook) {
+	phm.hooksMu.Lock()
+	defer phm.hooksMu.Unlock()
+	phm.hooks = append(phm.hooks, hook)
+}
+
+// RemoveHook removes a pool hook from the manager.
+func (phm *PoolHookManager) RemoveHook(hook PoolHook) {
+	phm.hooksMu.Lock()
+	defer phm.hooksMu.Unlock()
+
+	for i, h := range phm.hooks {
+		if h == hook {
+			// Remove hook by swapping with last element and truncating
+			phm.hooks[i] = phm.hooks[len(phm.hooks)-1]
+			phm.hooks = phm.hooks[:len(phm.hooks)-1]
+			break
+		}
+	}
+}
+
+// ProcessOnGet calls all OnGet hooks in order.
+// If any hook returns an error, processing stops and the error is returned.
+func (phm *PoolHookManager) ProcessOnGet(ctx context.Context, conn *Conn, isNewConn bool) error {
+	phm.hooksMu.RLock()
+	defer phm.hooksMu.RUnlock()
+
+	for _, hook := range phm.hooks {
+		if err := hook.OnGet(ctx, conn, isNewConn); err != nil {
+			return err
+		}
+	}
+	return nil
+}
+
+// ProcessOnPut calls all OnPut hooks in order.
+// The first hook that returns shouldRemove=true or shouldPool=false will stop processing.
+func (phm *PoolHookManager) ProcessOnPut(ctx context.Context, conn *Conn) (shouldPool bool, shouldRemove bool, err error) {
+	phm.hooksMu.RLock()
+	defer phm.hooksMu.RUnlock()
+
+	shouldPool = true // Default to pooling the connection
+
+	for _, hook := range phm.hooks {
+		hookShouldPool, hookShouldRemove, hookErr := hook.OnPut(ctx, conn)
+
+		if hookErr != nil {
+			return false, true, hookErr
+		}
+
+		// If any hook says to remove or not pool, respect that decision
+		if hookShouldRemove {
+			return false, true, nil
+		}
+
+		if !hookShouldPool {
+			shouldPool = false
+		}
+	}
+
+	return shouldPool, false, nil
+}
+
+// GetHookCount returns the number of registered hooks (for testing).
+func (phm *PoolHookManager) GetHookCount() int {
+	phm.hooksMu.RLock()
+	defer phm.hooksMu.RUnlock()
+	return len(phm.hooks)
+}
+
+// GetHooks returns a copy of all registered hooks.
+func (phm *PoolHookManager) GetHooks() []PoolHook {
+	phm.hooksMu.RLock()
+	defer phm.hooksMu.RUnlock()
+
+	hooks := make([]PoolHook, len(phm.hooks))
+	copy(hooks, phm.hooks)
+	return hooks
+}

internal/pool/hooks_test.go

@@ -0,0 +1,213 @@
+package pool
+
+import (
+	"context"
+	"errors"
+	"net"
+	"testing"
+	"time"
+)
+
+// TestHook for testing hook functionality
+type TestHook struct {
+	OnGetCalled  int
+	OnPutCalled  int
+	GetError     error
+	PutError     error
+	ShouldPool   bool
+	ShouldRemove bool
+}
+
+func (th *TestHook) OnGet(ctx context.Context, conn *Conn, isNewConn bool) error {
+	th.OnGetCalled++
+	return th.GetError
+}
+
+func (th *TestHook) OnPut(ctx context.Context, conn *Conn) (shouldPool bool, shouldRemove bool, err error) {
+	th.OnPutCalled++
+	return th.ShouldPool, th.ShouldRemove, th.PutError
+}
+
+func TestPoolHookManager(t *testing.T) {
+	manager := NewPoolHookManager()
+
+	// Test initial state
+	if manager.GetHookCount() != 0 {
+		t.Errorf("Expected 0 hooks initially, got %d", manager.GetHookCount())
+	}
+
+	// Add hooks
+	hook1 := &TestHook{ShouldPool: true}
+	hook2 := &TestHook{ShouldPool: true}
+
+	manager.AddHook(hook1)
+	manager.AddHook(hook2)
+
+	if manager.GetHookCount() != 2 {
+		t.Errorf("Expected 2 hooks after adding, got %d", manager.GetHookCount())
+	}
+
+	// Test ProcessOnGet
+	ctx := context.Background()
+	conn := &Conn{} // Mock connection
+
+	err := manager.ProcessOnGet(ctx, conn, false)
+	if err != nil {
+		t.Errorf("ProcessOnGet should not error: %v", err)
+	}
+
+	if hook1.OnGetCalled != 1 {
+		t.Errorf("Expected hook1.OnGetCalled to be 1, got %d", hook1.OnGetCalled)
+	}
+
+	if hook2.OnGetCalled != 1 {
+		t.Errorf("Expected hook2.OnGetCalled to be 1, got %d", hook2.OnGetCalled)
+	}
+
+	// Test ProcessOnPut
+	shouldPool, shouldRemove, err := manager.ProcessOnPut(ctx, conn)
+	if err != nil {
+		t.Errorf("ProcessOnPut should not error: %v", err)
+	}
+
+	if !shouldPool {
+		t.Error("Expected shouldPool to be true")
+	}
+
+	if shouldRemove {
+		t.Error("Expected shouldRemove to be false")
+	}
+
+	if hook1.OnPutCalled != 1 {
+		t.Errorf("Expected hook1.OnPutCalled to be 1, got %d", hook1.OnPutCalled)
+	}
+
+	if hook2.OnPutCalled != 1 {
+		t.Errorf("Expected hook2.OnPutCalled to be 1, got %d", hook2.OnPutCalled)
+	}
+
+	// Remove a hook
+	manager.RemoveHook(hook1)
+
+	if manager.GetHookCount() != 1 {
+		t.Errorf("Expected 1 hook after removing, got %d", manager.GetHookCount())
+	}
+}
+
+func TestHookErrorHandling(t *testing.T) {
+	manager := NewPoolHookManager()
+
+	// Hook that returns error on Get
+	errorHook := &TestHook{
+		GetError:   errors.New("test error"),
+		ShouldPool: true,
+	}
+
+	normalHook := &TestHook{ShouldPool: true}
+
+	manager.AddHook(errorHook)
+	manager.AddHook(normalHook)
+
+	ctx := context.Background()
+	conn := &Conn{}
+
+	// Test that error stops processing
+	err := manager.ProcessOnGet(ctx, conn, false)
+	if err == nil {
+		t.Error("Expected error from ProcessOnGet")
+	}
+
+	if errorHook.OnGetCalled != 1 {
+		t.Errorf("Expected errorHook.OnGetCalled to be 1, got %d", errorHook.OnGetCalled)
+	}
+
+	// normalHook should not be called due to error
+	if normalHook.OnGetCalled != 0 {
+		t.Errorf("Expected normalHook.OnGetCalled to be 0, got %d", normalHook.OnGetCalled)
+	}
+}
+
+func TestHookShouldRemove(t *testing.T) {
+	manager := NewPoolHookManager()
+
+	// Hook that says to remove connection
+	removeHook := &TestHook{
+		ShouldPool:   false,
+		ShouldRemove: true,
+	}
+
+	normalHook := &TestHook{ShouldPool: true}
+
+	manager.AddHook(removeHook)
+	manager.AddHook(normalHook)
+
+	ctx := context.Background()
+	conn := &Conn{}
+
+	shouldPool, shouldRemove, err := manager.ProcessOnPut(ctx, conn)
+	if err != nil {
+		t.Errorf("ProcessOnPut should not error: %v", err)
+	}
+
+	if shouldPool {
+		t.Error("Expected shouldPool to be false")
+	}
+
+	if !shouldRemove {
+		t.Error("Expected shouldRemove to be true")
+	}
+
+	if removeHook.OnPutCalled != 1 {
+		t.Errorf("Expected removeHook.OnPutCalled to be 1, got %d", removeHook.OnPutCalled)
+	}
+
+	// normalHook should not be called due to early return
+	if normalHook.OnPutCalled != 0 {
+		t.Errorf("Expected normalHook.OnPutCalled to be 0, got %d", normalHook.OnPutCalled)
+	}
+}
+
+func TestPoolWithHooks(t *testing.T) {
+	// Create a pool with hooks
+	hookManager := NewPoolHookManager()
+	testHook := &TestHook{ShouldPool: true}
+	hookManager.AddHook(testHook)
+
+	opt := &Options{
+		Dialer: func(ctx context.Context) (net.Conn, error) {
+			return &net.TCPConn{}, nil // Mock connection
+		},
+		PoolSize:    1,
+		DialTimeout: time.Second,
+	}
+
+	pool := NewConnPool(opt)
+	defer pool.Close()
+
+	// Add hook to pool after creation
+	pool.AddPoolHook(testHook)
+
+	// Verify hooks are initialized
+	if pool.hookManager == nil {
+		t.Error("Expected hookManager to be initialized")
+	}
+
+	if pool.hookManager.GetHookCount() != 1 {
+		t.Errorf("Expected 1 hook in pool, got %d", pool.hookManager.GetHookCount())
+	}
+
+	// Test adding hook to pool
+	additionalHook := &TestHook{ShouldPool: true}
+	pool.AddPoolHook(additionalHook)
+
+	if pool.hookManager.GetHookCount() != 2 {
+		t.Errorf("Expected 2 hooks after adding, got %d", pool.hookManager.GetHookCount())
+	}
+
+	// Test removing hook from pool
+	pool.RemovePoolHook(additionalHook)
+
+	if pool.hookManager.GetHookCount() != 1 {
+		t.Errorf("Expected 1 hook after removing, got %d", pool.hookManager.GetHookCount())
+	}
+}

internal/pool/pool.go

@@ -9,6 +9,8 @@ import (
 	"time"
 
 	"github.com/redis/go-redis/v9/internal"
+	"github.com/redis/go-redis/v9/internal/proto"
+	"github.com/redis/go-redis/v9/internal/util"
 )
 
 var (
@@ -21,6 +23,23 @@ var (
 
 	// ErrPoolTimeout timed out waiting to get a connection from the connection pool.
 	ErrPoolTimeout = errors.New("redis: connection pool timeout")
+
+	// popAttempts is the maximum number of attempts to find a usable connection
+	// when popping from the idle connection pool. This handles cases where connections
+	// are temporarily marked as unusable (e.g., during hitless upgrades or network issues).
+	// Value of 50 provides sufficient resilience without excessive overhead.
+	// This is capped by the idle connection count, so we won't loop excessively.
+	popAttempts = 50
+
+	// getAttempts is the maximum number of attempts to get a connection that passes
+	// hook validation (e.g., hitless upgrade hooks). This protects against race conditions
+	// where hooks might temporarily reject connections during cluster transitions.
+	// Value of 3 balances resilience with performance - most hook rejections resolve quickly.
+	getAttempts = 3
+
+	minTime      = time.Unix(-2208988800, 0) // Jan 1, 1900
+	maxTime      = minTime.Add(1<<63 - 1)
+	noExpiration = maxTime
 )
 
 var timers = sync.Pool{
@@ -37,11 +56,14 @@ type Stats struct {
 	Misses         uint32 // number of times free connection was NOT found in the pool
 	Timeouts       uint32 // number of times a wait timeout occurred
 	WaitCount      uint32 // number of times a connection was waited
+	Unusable       uint32 // number of times a connection was found to be unusable
 	WaitDurationNs int64  // total time spent for waiting a connection in nanoseconds
 
 	TotalConns uint32 // number of total connections in the pool
 	IdleConns  uint32 // number of idle connections in the pool
 	StaleConns uint32 // number of stale connections removed from the pool
+
+	PubSubStats PubSubStats
 }
 
 type Pooler interface {
@@ -56,24 +78,35 @@ type Pooler interface {
 	IdleLen() int
 	Stats() *Stats
 
+	AddPoolHook(hook PoolHook)
+	RemovePoolHook(hook PoolHook)
+
 	Close() error
 }
 
 type Options struct {
 	Dialer          func(context.Context) (net.Conn, error)
-
-	PoolFIFO        bool
-	PoolSize        int
-	DialTimeout     time.Duration
-	PoolTimeout     time.Duration
-	MinIdleConns    int
-	MaxIdleConns    int
-	MaxActiveConns  int
-	ConnMaxIdleTime time.Duration
-	ConnMaxLifetime time.Duration
-
 	ReadBufferSize  int
 	WriteBufferSize int
+
+	PoolFIFO                 bool
+	PoolSize                 int32
+	DialTimeout              time.Duration
+	PoolTimeout              time.Duration
+	MinIdleConns             int32
+	MaxIdleConns             int32
+	MaxActiveConns           int32
+	ConnMaxIdleTime          time.Duration
+	ConnMaxLifetime          time.Duration
+	PushNotificationsEnabled bool
+
+	// DialerRetries is the maximum number of retry attempts when dialing fails.
+	// Default: 5
+	DialerRetries int
+
+	// DialerRetryTimeout is the backoff duration between retry attempts.
+	// Default: 100ms
+	DialerRetryTimeout time.Duration
 }
 
 type lastDialErrorWrap struct {
@@ -89,16 +122,21 @@ type ConnPool struct {
 	queue chan struct{}
 
 	connsMu   sync.Mutex
-	conns     []*Conn
+	conns     map[uint64]*Conn
 	idleConns []*Conn
 
-	poolSize     int
-	idleConnsLen int
+	poolSize            atomic.Int32
+	idleConnsLen        atomic.Int32
+	idleCheckInProgress atomic.Bool
 
 	stats          Stats
 	waitDurationNs atomic.Int64
 
 	_closed uint32 // atomic
+
+	// Pool hooks manager for flexible connection processing
+	hookManagerMu sync.RWMutex
+	hookManager   *PoolHookManager
 }
 
 var _ Pooler = (*ConnPool)(nil)
@@ -108,34 +146,69 @@ func NewConnPool(opt *Options) *ConnPool {
 		cfg: opt,
 
 		queue:     make(chan struct{}, opt.PoolSize),
-		conns:     make([]*Conn, 0, opt.PoolSize),
+		conns:     make(map[uint64]*Conn),
 		idleConns: make([]*Conn, 0, opt.PoolSize),
 	}
 
+	// Only create MinIdleConns if explicitly requested (> 0)
+	// This avoids creating connections during pool initialization for tests
+	if opt.MinIdleConns > 0 {
 		p.connsMu.Lock()
 		p.checkMinIdleConns()
 		p.connsMu.Unlock()
+	}
 
 	return p
 }
 
+// initializeHooks sets up the pool hooks system.
+func (p *ConnPool) initializeHooks() {
+	p.hookManager = NewPoolHookManager()
+}
+
+// AddPoolHook adds a pool hook to the pool.
+func (p *ConnPool) AddPoolHook(hook PoolHook) {
+	p.hookManagerMu.Lock()
+	defer p.hookManagerMu.Unlock()
+
+	if p.hookManager == nil {
+		p.initializeHooks()
+	}
+	p.hookManager.AddHook(hook)
+}
+
+// RemovePoolHook removes a pool hook from the pool.
+func (p *ConnPool) RemovePoolHook(hook PoolHook) {
+	p.hookManagerMu.Lock()
+	defer p.hookManagerMu.Unlock()
+
+	if p.hookManager != nil {
+		p.hookManager.RemoveHook(hook)
+	}
+}
+
 func (p *ConnPool) checkMinIdleConns() {
+	if !p.idleCheckInProgress.CompareAndSwap(false, true) {
+		return
+	}
+	defer p.idleCheckInProgress.Store(false)
+
 	if p.cfg.MinIdleConns == 0 {
 		return
 	}
-	for p.poolSize < p.cfg.PoolSize && p.idleConnsLen < p.cfg.MinIdleConns {
+
+	// Only create idle connections if we haven't reached the total pool size limit
+	// MinIdleConns should be a subset of PoolSize, not additional connections
+	for p.poolSize.Load() < p.cfg.PoolSize && p.idleConnsLen.Load() < p.cfg.MinIdleConns {
 		select {
 		case p.queue <- struct{}{}:
-			p.poolSize++
-			p.idleConnsLen++
-
+			p.poolSize.Add(1)
+			p.idleConnsLen.Add(1)
 			go func() {
 				defer func() {
 					if err := recover(); err != nil {
-						p.connsMu.Lock()
-						p.poolSize--
-						p.idleConnsLen--
-						p.connsMu.Unlock()
+						p.poolSize.Add(-1)
+						p.idleConnsLen.Add(-1)
 
 						p.freeTurn()
 						internal.Logger.Printf(context.Background(), "addIdleConn panic: %+v", err)
@@ -144,12 +217,9 @@ func (p *ConnPool) checkMinIdleConns() {
 
 				err := p.addIdleConn()
 				if err != nil && err != ErrClosed {
-					p.connsMu.Lock()
-					p.poolSize--
-					p.idleConnsLen--
-					p.connsMu.Unlock()
+					p.poolSize.Add(-1)
+					p.idleConnsLen.Add(-1)
 				}
-
 				p.freeTurn()
 			}()
 		default:
@@ -166,6 +236,9 @@ func (p *ConnPool) addIdleConn() error {
 	if err != nil {
 		return err
 	}
+	// Mark connection as usable after successful creation
+	// This is essential for normal pool operations
+	cn.SetUsable(true)
 
 	p.connsMu.Lock()
 	defer p.connsMu.Unlock()
@@ -176,11 +249,15 @@ func (p *ConnPool) addIdleConn() error {
 		return ErrClosed
 	}
 
-	p.conns = append(p.conns, cn)
+	p.conns[cn.GetID()] = cn
 	p.idleConns = append(p.idleConns, cn)
 	return nil
 }
 
+// NewConn creates a new connection and returns it to the user.
+// This will still obey MaxActiveConns but will not include it in the pool and won't increase the pool size.
+//
+// NOTE: If you directly get a connection from the pool, it won't be pooled and won't support hitless upgrades.
 func (p *ConnPool) NewConn(ctx context.Context) (*Conn, error) {
 	return p.newConn(ctx, false)
 }
@@ -190,33 +267,44 @@ func (p *ConnPool) newConn(ctx context.Context, pooled bool) (*Conn, error) {
 		return nil, ErrClosed
 	}
 
-	p.connsMu.Lock()
-	if p.cfg.MaxActiveConns > 0 && p.poolSize >= p.cfg.MaxActiveConns {
-		p.connsMu.Unlock()
+	if p.cfg.MaxActiveConns > 0 && p.poolSize.Load() >= int32(p.cfg.MaxActiveConns) {
 		return nil, ErrPoolExhausted
 	}
-	p.connsMu.Unlock()
 
-	cn, err := p.dialConn(ctx, pooled)
+	dialCtx, cancel := context.WithTimeout(ctx, p.cfg.DialTimeout)
+	defer cancel()
+	cn, err := p.dialConn(dialCtx, pooled)
 	if err != nil {
 		return nil, err
 	}
+	// Mark connection as usable after successful creation
+	// This is essential for normal pool operations
+	cn.SetUsable(true)
 
-	p.connsMu.Lock()
-	defer p.connsMu.Unlock()
-
-	if p.cfg.MaxActiveConns > 0 && p.poolSize >= p.cfg.MaxActiveConns {
+	if p.cfg.MaxActiveConns > 0 && p.poolSize.Load() > int32(p.cfg.MaxActiveConns) {
 		_ = cn.Close()
 		return nil, ErrPoolExhausted
 	}
 
-	p.conns = append(p.conns, cn)
+	p.connsMu.Lock()
+	defer p.connsMu.Unlock()
+	if p.closed() {
+		_ = cn.Close()
+		return nil, ErrClosed
+	}
+	// Check if pool was closed while we were waiting for the lock
+	if p.conns == nil {
+		p.conns = make(map[uint64]*Conn)
+	}
+	p.conns[cn.GetID()] = cn
+
 	if pooled {
 		// If pool is full remove the cn on next Put.
-		if p.poolSize >= p.cfg.PoolSize {
+		currentPoolSize := p.poolSize.Load()
+		if currentPoolSize >= p.cfg.PoolSize {
 			cn.pooled = false
 		} else {
-			p.poolSize++
+			p.poolSize.Add(1)
 		}
 	}
 
@@ -232,18 +320,57 @@ func (p *ConnPool) dialConn(ctx context.Context, pooled bool) (*Conn, error) {
 		return nil, p.getLastDialError()
 	}
 
+	// Retry dialing with backoff
+	// the context timeout is already handled by the context passed in
+	// so we may never reach the max retries, higher values don't hurt
+	maxRetries := p.cfg.DialerRetries
+	if maxRetries <= 0 {
+		maxRetries = 5 // Default value
+	}
+	backoffDuration := p.cfg.DialerRetryTimeout
+	if backoffDuration <= 0 {
+		backoffDuration = 100 * time.Millisecond // Default value
+	}
+
+	var lastErr error
+	shouldLoop := true
+	// when the timeout is reached, we should stop retrying
+	// but keep the lastErr to return to the caller
+	// instead of a generic context deadline exceeded error
+	for attempt := 0; (attempt < maxRetries) && shouldLoop; attempt++ {
 		netConn, err := p.cfg.Dialer(ctx)
 		if err != nil {
-		p.setLastDialError(err)
+			lastErr = err
+			// Add backoff delay for retry attempts
+			// (not for the first attempt, do at least one)
+			select {
+			case <-ctx.Done():
+				shouldLoop = false
+			case <-time.After(backoffDuration):
+				// Continue with retry
+			}
+			continue
+		}
+
+		// Success - create connection
+		cn := NewConnWithBufferSize(netConn, p.cfg.ReadBufferSize, p.cfg.WriteBufferSize)
+		cn.pooled = pooled
+		if p.cfg.ConnMaxLifetime > 0 {
+			cn.expiresAt = time.Now().Add(p.cfg.ConnMaxLifetime)
+		} else {
+			cn.expiresAt = noExpiration
+		}
+
+		return cn, nil
+	}
+
+	internal.Logger.Printf(ctx, "redis: connection pool: failed to dial after %d attempts: %v", maxRetries, lastErr)
+	// All retries failed - handle error tracking
+	p.setLastDialError(lastErr)
 	if atomic.AddUint32(&p.dialErrorsNum, 1) == uint32(p.cfg.PoolSize) {
 		go p.tryDial()
 	}
-		return nil, err
-	}
-
-	cn := NewConnWithBufferSize(netConn, p.cfg.ReadBufferSize, p.cfg.WriteBufferSize)
-	cn.pooled = pooled
-	return cn, nil
+	return nil, lastErr
 }
 
 func (p *ConnPool) tryDial() {
@@ -283,6 +410,14 @@ func (p *ConnPool) getLastDialError() error {
 
 // Get returns existed connection from the pool or creates a new one.
 func (p *ConnPool) Get(ctx context.Context) (*Conn, error) {
+	return p.getConn(ctx)
+}
+
+// getConn returns a connection from the pool.
+func (p *ConnPool) getConn(ctx context.Context) (*Conn, error) {
+	var cn *Conn
+	var err error
+
 	if p.closed() {
 		return nil, ErrClosed
 	}
@@ -291,9 +426,17 @@ func (p *ConnPool) Get(ctx context.Context) (*Conn, error) {
 		return nil, err
 	}
 
+	now := time.Now()
+	attempts := 0
 	for {
+		if attempts >= getAttempts {
+			internal.Logger.Printf(ctx, "redis: connection pool: was not able to get a healthy connection after %d attempts", attempts)
+			break
+		}
+		attempts++
+
 		p.connsMu.Lock()
-		cn, err := p.popIdle()
+		cn, err = p.popIdle()
 		p.connsMu.Unlock()
 
 		if err != nil {
@@ -305,11 +448,25 @@ func (p *ConnPool) Get(ctx context.Context) (*Conn, error) {
 			break
 		}
 
-		if !p.isHealthyConn(cn) {
+		if !p.isHealthyConn(cn, now) {
 			_ = p.CloseConn(cn)
 			continue
 		}
 
+		// Process connection using the hooks system
+		p.hookManagerMu.RLock()
+		hookManager := p.hookManager
+		p.hookManagerMu.RUnlock()
+
+		if hookManager != nil {
+			if err := hookManager.ProcessOnGet(ctx, cn, false); err != nil {
+				internal.Logger.Printf(ctx, "redis: connection pool: failed to process idle connection by hook: %v", err)
+				// Failed to process connection, discard it
+				_ = p.CloseConn(cn)
+				continue
+			}
+		}
+
 		atomic.AddUint32(&p.stats.Hits, 1)
 		return cn, nil
 	}
@@ -322,6 +479,19 @@ func (p *ConnPool) Get(ctx context.Context) (*Conn, error) {
 		return nil, err
 	}
 
+	// Process connection using the hooks system
+	p.hookManagerMu.RLock()
+	hookManager := p.hookManager
+	p.hookManagerMu.RUnlock()
+
+	if hookManager != nil {
+		if err := hookManager.ProcessOnGet(ctx, newcn, true); err != nil {
+			// Failed to process connection, discard it
+			internal.Logger.Printf(ctx, "redis: connection pool: failed to process new connection by hook: %v", err)
+			_ = p.CloseConn(newcn)
+			return nil, err
+		}
+	}
 	return newcn, nil
 }
 
@@ -350,7 +520,7 @@ func (p *ConnPool) waitTurn(ctx context.Context) error {
 		}
 		return ctx.Err()
 	case p.queue <- struct{}{}:
-		p.waitDurationNs.Add(time.Since(start).Nanoseconds())
+		p.waitDurationNs.Add(time.Now().UnixNano() - start.UnixNano())
 		atomic.AddUint32(&p.stats.WaitCount, 1)
 		if !timer.Stop() {
 			<-timer.C
@@ -370,52 +540,130 @@ func (p *ConnPool) popIdle() (*Conn, error) {
 	if p.closed() {
 		return nil, ErrClosed
 	}
+	defer p.checkMinIdleConns()
+
 	n := len(p.idleConns)
 	if n == 0 {
 		return nil, nil
 	}
 
 	var cn *Conn
+	attempts := 0
+
+	maxAttempts := util.Min(popAttempts, n)
+	for attempts < maxAttempts {
+		if len(p.idleConns) == 0 {
+			return nil, nil
+		}
+
 		if p.cfg.PoolFIFO {
 			cn = p.idleConns[0]
 			copy(p.idleConns, p.idleConns[1:])
-		p.idleConns = p.idleConns[:n-1]
+			p.idleConns = p.idleConns[:len(p.idleConns)-1]
 		} else {
-		idx := n - 1
+			idx := len(p.idleConns) - 1
 			cn = p.idleConns[idx]
 			p.idleConns = p.idleConns[:idx]
 		}
-	p.idleConnsLen--
-	p.checkMinIdleConns()
+		attempts++
+
+		if cn.IsUsable() {
+			p.idleConnsLen.Add(-1)
+			break
+		}
+
+		// Connection is not usable, put it back in the pool
+		if p.cfg.PoolFIFO {
+			// FIFO: put at end (will be picked up last since we pop from front)
+			p.idleConns = append(p.idleConns, cn)
+		} else {
+			// LIFO: put at beginning (will be picked up last since we pop from end)
+			p.idleConns = append([]*Conn{cn}, p.idleConns...)
+		}
+		cn = nil
+	}
+
+	// If we exhausted all attempts without finding a usable connection, return nil
+	if attempts > 1 && attempts >= maxAttempts && int32(attempts) >= p.poolSize.Load() {
+		internal.Logger.Printf(context.Background(), "redis: connection pool: failed to get a usable connection after %d attempts", attempts)
+		return nil, nil
+	}
+
 	return cn, nil
 }
 
 func (p *ConnPool) Put(ctx context.Context, cn *Conn) {
-	if cn.rd.Buffered() > 0 {
-		internal.Logger.Printf(ctx, "Conn has unread data")
-		p.Remove(ctx, cn, BadConnError{})
+	// Process connection using the hooks system
+	shouldPool := true
+	shouldRemove := false
+	var err error
+
+	if cn.HasBufferedData() {
+		// Peek at the reply type to check if it's a push notification
+		if replyType, err := cn.PeekReplyTypeSafe(); err != nil || replyType != proto.RespPush {
+			// Not a push notification or error peeking, remove connection
+			internal.Logger.Printf(ctx, "Conn has unread data (not push notification), removing it")
+			p.Remove(ctx, cn, err)
+		}
+		// It's a push notification, allow pooling (client will handle it)
+	}
+
+	p.hookManagerMu.RLock()
+	hookManager := p.hookManager
+	p.hookManagerMu.RUnlock()
+
+	if hookManager != nil {
+		shouldPool, shouldRemove, err = hookManager.ProcessOnPut(ctx, cn)
+		if err != nil {
+			internal.Logger.Printf(ctx, "Connection hook error: %v", err)
+			p.Remove(ctx, cn, err)
+			return
+		}
+	}
+
+	// If hooks say to remove the connection, do so
+	if shouldRemove {
+		p.Remove(ctx, cn, errors.New("hook requested removal"))
+		return
+	}
+
+	// If processor says not to pool the connection, remove it
+	if !shouldPool {
+		p.Remove(ctx, cn, errors.New("hook requested no pooling"))
 		return
 	}
 
 	if !cn.pooled {
-		p.Remove(ctx, cn, nil)
+		p.Remove(ctx, cn, errors.New("connection not pooled"))
 		return
 	}
 
 	var shouldCloseConn bool
 
+	if p.cfg.MaxIdleConns == 0 || p.idleConnsLen.Load() < p.cfg.MaxIdleConns {
+		// unusable conns are expected to become usable at some point (background process is reconnecting them)
+		// put them at the opposite end of the queue
+		if !cn.IsUsable() {
+			if p.cfg.PoolFIFO {
 				p.connsMu.Lock()
-
-	if p.cfg.MaxIdleConns == 0 || p.idleConnsLen < p.cfg.MaxIdleConns {
 				p.idleConns = append(p.idleConns, cn)
-		p.idleConnsLen++
+				p.connsMu.Unlock()
 			} else {
-		p.removeConn(cn)
+				p.connsMu.Lock()
+				p.idleConns = append([]*Conn{cn}, p.idleConns...)
+				p.connsMu.Unlock()
+			}
+		} else {
+			p.connsMu.Lock()
+			p.idleConns = append(p.idleConns, cn)
+			p.connsMu.Unlock()
+		}
+		p.idleConnsLen.Add(1)
+	} else {
+		p.removeConnWithLock(cn)
 		shouldCloseConn = true
 	}
 
-	p.connsMu.Unlock()
-
 	p.freeTurn()
 
 	if shouldCloseConn {
@@ -425,8 +673,13 @@ func (p *ConnPool) Put(ctx context.Context, cn *Conn) {
 
 func (p *ConnPool) Remove(_ context.Context, cn *Conn, reason error) {
 	p.removeConnWithLock(cn)
+
 	p.freeTurn()
+
 	_ = p.closeConn(cn)
+
+	// Check if we need to create new idle connections to maintain MinIdleConns
+	p.checkMinIdleConns()
 }
 
 func (p *ConnPool) CloseConn(cn *Conn) error {
@@ -441,17 +694,23 @@ func (p *ConnPool) removeConnWithLock(cn *Conn) {
 }
 
 func (p *ConnPool) removeConn(cn *Conn) {
-	for i, c := range p.conns {
-		if c == cn {
-			p.conns = append(p.conns[:i], p.conns[i+1:]...)
+	cid := cn.GetID()
+	delete(p.conns, cid)
+	atomic.AddUint32(&p.stats.StaleConns, 1)
+
+	// Decrement pool size counter when removing a connection
 	if cn.pooled {
-				p.poolSize--
-				p.checkMinIdleConns()
-			}
+		p.poolSize.Add(-1)
+		// this can be idle conn
+		for idx, ic := range p.idleConns {
+			if ic.GetID() == cid {
+				internal.Logger.Printf(context.Background(), "redis: connection pool: removing idle conn[%d]", cid)
+				p.idleConns = append(p.idleConns[:idx], p.idleConns[idx+1:]...)
+				p.idleConnsLen.Add(-1)
 				break
 			}
 		}
-	atomic.AddUint32(&p.stats.StaleConns, 1)
+	}
 }
 
 func (p *ConnPool) closeConn(cn *Conn) error {
@@ -469,9 +728,9 @@ func (p *ConnPool) Len() int {
 // IdleLen returns number of idle connections.
 func (p *ConnPool) IdleLen() int {
 	p.connsMu.Lock()
-	n := p.idleConnsLen
+	n := p.idleConnsLen.Load()
 	p.connsMu.Unlock()
-	return n
+	return int(n)
 }
 
 func (p *ConnPool) Stats() *Stats {
@@ -480,6 +739,7 @@ func (p *ConnPool) Stats() *Stats {
 		Misses:         atomic.LoadUint32(&p.stats.Misses),
 		Timeouts:       atomic.LoadUint32(&p.stats.Timeouts),
 		WaitCount:      atomic.LoadUint32(&p.stats.WaitCount),
+		Unusable:       atomic.LoadUint32(&p.stats.Unusable),
 		WaitDurationNs: p.waitDurationNs.Load(),
 
 		TotalConns: uint32(p.Len()),
@@ -520,28 +780,45 @@ func (p *ConnPool) Close() error {
 		}
 	}
 	p.conns = nil
-	p.poolSize = 0
+	p.poolSize.Store(0)
 	p.idleConns = nil
-	p.idleConnsLen = 0
+	p.idleConnsLen.Store(0)
 	p.connsMu.Unlock()
 
 	return firstErr
 }
 
-func (p *ConnPool) isHealthyConn(cn *Conn) bool {
-	now := time.Now()
-
-	if p.cfg.ConnMaxLifetime > 0 && now.Sub(cn.createdAt) >= p.cfg.ConnMaxLifetime {
+func (p *ConnPool) isHealthyConn(cn *Conn, now time.Time) bool {
+	// slight optimization, check expiresAt first.
+	if cn.expiresAt.Before(now) {
 		return false
 	}
+
+	// Check if connection has exceeded idle timeout
 	if p.cfg.ConnMaxIdleTime > 0 && now.Sub(cn.UsedAt()) >= p.cfg.ConnMaxIdleTime {
 		return false
 	}
 
-	if connCheck(cn.netConn) != nil {
+	cn.SetUsedAt(now)
+	// Check basic connection health
+	// Use GetNetConn() to safely access netConn and avoid data races
+	if err := connCheck(cn.getNetConn()); err != nil {
+		// If there's unexpected data, it might be push notifications (RESP3)
+		// However, push notification processing is now handled by the client
+		// before WithReader to ensure proper context is available to handlers
+		if p.cfg.PushNotificationsEnabled && err == errUnexpectedRead {
+			// we know that there is something in the buffer, so peek at the next reply type without
+			// the potential to block
+			if replyType, err := cn.rd.PeekReplyType(); err == nil && replyType == proto.RespPush {
+				// For RESP3 connections with push notifications, we allow some buffered data
+				// The client will process these notifications before using the connection
+				internal.Logger.Printf(context.Background(), "push: connection has buffered data, likely push notifications - will be processed by client")
+				return true // Connection is healthy, client will handle notifications
+			}
+			return false // Unexpected data, not push notifications, connection is unhealthy
+		} else {
 			return false
 		}
-
-	cn.SetUsedAt(now)
+	}
 	return true
 }

internal/pool/pool_single.go

@@ -1,6 +1,8 @@
 package pool
 
-import "context"
+import (
+	"context"
+)
 
 type SingleConnPool struct {
 	pool      Pooler
@@ -56,3 +58,7 @@ func (p *SingleConnPool) IdleLen() int {
 func (p *SingleConnPool) Stats() *Stats {
 	return &Stats{}
 }
+
+func (p *SingleConnPool) AddPoolHook(hook PoolHook) {}
+
+func (p *SingleConnPool) RemovePoolHook(hook PoolHook) {}

internal/pool/pool_sticky.go

@@ -199,3 +199,7 @@ func (p *StickyConnPool) IdleLen() int {
 func (p *StickyConnPool) Stats() *Stats {
 	return &Stats{}
 }
+
+func (p *StickyConnPool) AddPoolHook(hook PoolHook) {}
+
+func (p *StickyConnPool) RemovePoolHook(hook PoolHook) {}

internal/pool/pool_test.go

@@ -2,15 +2,17 @@ package pool_test
 
 import (
 	"context"
+	"errors"
 	"net"
 	"sync"
+	"sync/atomic"
 	"testing"
 	"time"
 
 	. "github.com/bsm/ginkgo/v2"
 	. "github.com/bsm/gomega"
-
 	"github.com/redis/go-redis/v9/internal/pool"
+	"github.com/redis/go-redis/v9/logging"
 )
 
 var _ = Describe("ConnPool", func() {
@@ -20,7 +22,7 @@ var _ = Describe("ConnPool", func() {
 	BeforeEach(func() {
 		connPool = pool.NewConnPool(&pool.Options{
 			Dialer:          dummyDialer,
-			PoolSize:        10,
+			PoolSize:        int32(10),
 			PoolTimeout:     time.Hour,
 			DialTimeout:     1 * time.Second,
 			ConnMaxIdleTime: time.Millisecond,
@@ -45,11 +47,11 @@ var _ = Describe("ConnPool", func() {
 				<-closedChan
 				return &net.TCPConn{}, nil
 			},
-			PoolSize:        10,
+			PoolSize:        int32(10),
 			PoolTimeout:     time.Hour,
 			DialTimeout:     1 * time.Second,
 			ConnMaxIdleTime: time.Millisecond,
-			MinIdleConns:    minIdleConns,
+			MinIdleConns:    int32(minIdleConns),
 		})
 		wg.Wait()
 		Expect(connPool.Close()).NotTo(HaveOccurred())
@@ -105,7 +107,7 @@ var _ = Describe("ConnPool", func() {
 			// ok
 		}
 
-		connPool.Remove(ctx, cn, nil)
+		connPool.Remove(ctx, cn, errors.New("test"))
 
 		// Check that Get is unblocked.
 		select {
@@ -130,8 +132,8 @@ var _ = Describe("MinIdleConns", func() {
 	newConnPool := func() *pool.ConnPool {
 		connPool := pool.NewConnPool(&pool.Options{
 			Dialer:          dummyDialer,
-			PoolSize:        poolSize,
-			MinIdleConns:    minIdleConns,
+			PoolSize:        int32(poolSize),
+			MinIdleConns:    int32(minIdleConns),
 			PoolTimeout:     100 * time.Millisecond,
 			DialTimeout:     1 * time.Second,
 			ConnMaxIdleTime: -1,
@@ -168,7 +170,7 @@ var _ = Describe("MinIdleConns", func() {
 
 			Context("after Remove", func() {
 				BeforeEach(func() {
-					connPool.Remove(ctx, cn, nil)
+					connPool.Remove(ctx, cn, errors.New("test"))
 				})
 
 				It("has idle connections", func() {
@@ -245,7 +247,7 @@ var _ = Describe("MinIdleConns", func() {
 				BeforeEach(func() {
 					perform(len(cns), func(i int) {
 						mu.RLock()
-						connPool.Remove(ctx, cns[i], nil)
+						connPool.Remove(ctx, cns[i], errors.New("test"))
 						mu.RUnlock()
 					})
 
@@ -309,7 +311,7 @@ var _ = Describe("race", func() {
 	It("does not happen on Get, Put, and Remove", func() {
 		connPool = pool.NewConnPool(&pool.Options{
 			Dialer:          dummyDialer,
-			PoolSize:        10,
+			PoolSize:        int32(10),
 			PoolTimeout:     time.Minute,
 			DialTimeout:     1 * time.Second,
 			ConnMaxIdleTime: time.Millisecond,
@@ -328,7 +330,7 @@ var _ = Describe("race", func() {
 				cn, err := connPool.Get(ctx)
 				Expect(err).NotTo(HaveOccurred())
 				if err == nil {
-					connPool.Remove(ctx, cn, nil)
+					connPool.Remove(ctx, cn, errors.New("test"))
 				}
 			}
 		})
@@ -339,15 +341,15 @@ var _ = Describe("race", func() {
 			Dialer: func(ctx context.Context) (net.Conn, error) {
 				return &net.TCPConn{}, nil
 			},
-			PoolSize:     1000,
-			MinIdleConns: 50,
+			PoolSize:     int32(1000),
+			MinIdleConns: int32(50),
 			PoolTimeout:  3 * time.Second,
 			DialTimeout:  1 * time.Second,
 		}
 		p := pool.NewConnPool(opt)
 
 		var wg sync.WaitGroup
-		for i := 0; i < opt.PoolSize; i++ {
+		for i := int32(0); i < opt.PoolSize; i++ {
 			wg.Add(1)
 			go func() {
 				defer wg.Done()
@@ -366,8 +368,8 @@ var _ = Describe("race", func() {
 			Dialer: func(ctx context.Context) (net.Conn, error) {
 				panic("test panic")
 			},
-			PoolSize:     100,
-			MinIdleConns: 30,
+			PoolSize:     int32(100),
+			MinIdleConns: int32(30),
 		}
 		p := pool.NewConnPool(opt)
 
@@ -384,7 +386,7 @@ var _ = Describe("race", func() {
 			Dialer: func(ctx context.Context) (net.Conn, error) {
 				return &net.TCPConn{}, nil
 			},
-			PoolSize:    1,
+			PoolSize:    int32(1),
 			PoolTimeout: 3 * time.Second,
 		}
 		p := pool.NewConnPool(opt)
@@ -415,7 +417,7 @@ var _ = Describe("race", func() {
 
 				return &net.TCPConn{}, nil
 			},
-			PoolSize:    1,
+			PoolSize:    int32(1),
 			PoolTimeout: testPoolTimeout,
 		}
 		p := pool.NewConnPool(opt)
@@ -435,3 +437,73 @@ var _ = Describe("race", func() {
 		Expect(stats.Timeouts).To(Equal(uint32(1)))
 	})
 })
+
+// TestDialerRetryConfiguration tests the new DialerRetries and DialerRetryTimeout options
+func TestDialerRetryConfiguration(t *testing.T) {
+	ctx := context.Background()
+
+	t.Run("CustomDialerRetries", func(t *testing.T) {
+		var attempts int64
+		failingDialer := func(ctx context.Context) (net.Conn, error) {
+			atomic.AddInt64(&attempts, 1)
+			return nil, errors.New("dial failed")
+		}
+
+		connPool := pool.NewConnPool(&pool.Options{
+			Dialer:             failingDialer,
+			PoolSize:           1,
+			PoolTimeout:        time.Second,
+			DialTimeout:        time.Second,
+			DialerRetries:      3,                     // Custom retry count
+			DialerRetryTimeout: 10 * time.Millisecond, // Fast retries for testing
+		})
+		defer connPool.Close()
+
+		_, err := connPool.Get(ctx)
+		if err == nil {
+			t.Error("Expected error from failing dialer")
+		}
+
+		// Should have attempted at least 3 times (DialerRetries = 3)
+		// There might be additional attempts due to pool logic
+		finalAttempts := atomic.LoadInt64(&attempts)
+		if finalAttempts < 3 {
+			t.Errorf("Expected at least 3 dial attempts, got %d", finalAttempts)
+		}
+		if finalAttempts > 6 {
+			t.Errorf("Expected around 3 dial attempts, got %d (too many)", finalAttempts)
+		}
+	})
+
+	t.Run("DefaultDialerRetries", func(t *testing.T) {
+		var attempts int64
+		failingDialer := func(ctx context.Context) (net.Conn, error) {
+			atomic.AddInt64(&attempts, 1)
+			return nil, errors.New("dial failed")
+		}
+
+		connPool := pool.NewConnPool(&pool.Options{
+			Dialer:      failingDialer,
+			PoolSize:    1,
+			PoolTimeout: time.Second,
+			DialTimeout: time.Second,
+			// DialerRetries and DialerRetryTimeout not set - should use defaults
+		})
+		defer connPool.Close()
+
+		_, err := connPool.Get(ctx)
+		if err == nil {
+			t.Error("Expected error from failing dialer")
+		}
+
+		// Should have attempted 5 times (default DialerRetries = 5)
+		finalAttempts := atomic.LoadInt64(&attempts)
+		if finalAttempts != 5 {
+			t.Errorf("Expected 5 dial attempts (default), got %d", finalAttempts)
+		}
+	})
+}
+
+func init() {
+	logging.Disable()
+}

internal/pool/pubsub.go

@@ -0,0 +1,78 @@
+package pool
+
+import (
+	"context"
+	"net"
+	"sync"
+	"sync/atomic"
+)
+
+type PubSubStats struct {
+	Created   uint32
+	Untracked uint32
+	Active    uint32
+}
+
+// PubSubPool manages a pool of PubSub connections.
+type PubSubPool struct {
+	opt       *Options
+	netDialer func(ctx context.Context, network, addr string) (net.Conn, error)
+
+	// Map to track active PubSub connections
+	activeConns sync.Map // map[uint64]*Conn (connID -> conn)
+	closed      atomic.Bool
+	stats       PubSubStats
+}
+
+func NewPubSubPool(opt *Options, netDialer func(ctx context.Context, network, addr string) (net.Conn, error)) *PubSubPool {
+	return &PubSubPool{
+		opt:       opt,
+		netDialer: netDialer,
+	}
+}
+
+func (p *PubSubPool) NewConn(ctx context.Context, network string, addr string, channels []string) (*Conn, error) {
+	if p.closed.Load() {
+		return nil, ErrClosed
+	}
+
+	netConn, err := p.netDialer(ctx, network, addr)
+	if err != nil {
+		return nil, err
+	}
+	cn := NewConnWithBufferSize(netConn, p.opt.ReadBufferSize, p.opt.WriteBufferSize)
+	cn.pubsub = true
+	atomic.AddUint32(&p.stats.Created, 1)
+	return cn, nil
+
+}
+
+func (p *PubSubPool) TrackConn(cn *Conn) {
+	atomic.AddUint32(&p.stats.Active, 1)
+	p.activeConns.Store(cn.GetID(), cn)
+}
+
+func (p *PubSubPool) UntrackConn(cn *Conn) {
+	atomic.AddUint32(&p.stats.Active, ^uint32(0))
+	atomic.AddUint32(&p.stats.Untracked, 1)
+	p.activeConns.Delete(cn.GetID())
+}
+
+func (p *PubSubPool) Close() error {
+	p.closed.Store(true)
+	p.activeConns.Range(func(key, value interface{}) bool {
+		cn := value.(*Conn)
+		_ = cn.Close()
+		return true
+	})
+	return nil
+}
+
+func (p *PubSubPool) Stats() *PubSubStats {
+	// load stats atomically
+	return &PubSubStats{
+		Created:   atomic.LoadUint32(&p.stats.Created),
+		Untracked: atomic.LoadUint32(&p.stats.Untracked),
+		Active:    atomic.LoadUint32(&p.stats.Active),
+	}
+}

internal/proto/peek_push_notification_test.go

@@ -0,0 +1,614 @@
+package proto
+
+import (
+	"bytes"
+	"fmt"
+	"math/rand"
+	"strings"
+	"testing"
+)
+
+// TestPeekPushNotificationName tests the updated PeekPushNotificationName method
+func TestPeekPushNotificationName(t *testing.T) {
+	t.Run("ValidPushNotifications", func(t *testing.T) {
+		testCases := []struct {
+			name         string
+			notification string
+			expected     string
+		}{
+			{"MOVING", "MOVING", "MOVING"},
+			{"MIGRATING", "MIGRATING", "MIGRATING"},
+			{"MIGRATED", "MIGRATED", "MIGRATED"},
+			{"FAILING_OVER", "FAILING_OVER", "FAILING_OVER"},
+			{"FAILED_OVER", "FAILED_OVER", "FAILED_OVER"},
+			{"message", "message", "message"},
+			{"pmessage", "pmessage", "pmessage"},
+			{"subscribe", "subscribe", "subscribe"},
+			{"unsubscribe", "unsubscribe", "unsubscribe"},
+			{"psubscribe", "psubscribe", "psubscribe"},
+			{"punsubscribe", "punsubscribe", "punsubscribe"},
+			{"smessage", "smessage", "smessage"},
+			{"ssubscribe", "ssubscribe", "ssubscribe"},
+			{"sunsubscribe", "sunsubscribe", "sunsubscribe"},
+			{"custom", "custom", "custom"},
+			{"short", "a", "a"},
+			{"empty", "", ""},
+		}
+
+		for _, tc := range testCases {
+			t.Run(tc.name, func(t *testing.T) {
+				buf := createValidPushNotification(tc.notification, "data")
+				reader := NewReader(buf)
+
+				// Prime the buffer by peeking first
+				_, _ = reader.rd.Peek(1)
+
+				name, err := reader.PeekPushNotificationName()
+				if err != nil {
+					t.Errorf("PeekPushNotificationName should not error for valid notification: %v", err)
+				}
+
+				if name != tc.expected {
+					t.Errorf("Expected notification name '%s', got '%s'", tc.expected, name)
+				}
+			})
+		}
+	})
+
+	t.Run("NotificationWithMultipleArguments", func(t *testing.T) {
+		// Create push notification with multiple arguments
+		buf := createPushNotificationWithArgs("MOVING", "slot", "123", "from", "node1", "to", "node2")
+		reader := NewReader(buf)
+
+		// Prime the buffer
+		_, _ = reader.rd.Peek(1)
+
+		name, err := reader.PeekPushNotificationName()
+		if err != nil {
+			t.Errorf("PeekPushNotificationName should not error: %v", err)
+		}
+
+		if name != "MOVING" {
+			t.Errorf("Expected 'MOVING', got '%s'", name)
+		}
+	})
+
+	t.Run("SingleElementNotification", func(t *testing.T) {
+		// Create push notification with single element
+		buf := createSingleElementPushNotification("TEST")
+		reader := NewReader(buf)
+
+		// Prime the buffer
+		_, _ = reader.rd.Peek(1)
+
+		name, err := reader.PeekPushNotificationName()
+		if err != nil {
+			t.Errorf("PeekPushNotificationName should not error: %v", err)
+		}
+
+		if name != "TEST" {
+			t.Errorf("Expected 'TEST', got '%s'", name)
+		}
+	})
+
+	t.Run("ErrorDetection", func(t *testing.T) {
+		t.Run("NotPushNotification", func(t *testing.T) {
+			// Test with regular array instead of push notification
+			buf := &bytes.Buffer{}
+			buf.WriteString("*2\r\n$6\r\nMOVING\r\n$4\r\ndata\r\n")
+			reader := NewReader(buf)
+
+			_, err := reader.PeekPushNotificationName()
+			if err == nil {
+				t.Error("PeekPushNotificationName should error for non-push notification")
+			}
+
+			// The error might be "no data available" or "can't parse push notification"
+			if !strings.Contains(err.Error(), "can't peek push notification name") {
+				t.Errorf("Error should mention push notification parsing, got: %v", err)
+			}
+		})
+
+		t.Run("InsufficientData", func(t *testing.T) {
+			// Test with buffer smaller than peek size - this might panic due to bounds checking
+			buf := &bytes.Buffer{}
+			buf.WriteString(">")
+			reader := NewReader(buf)
+
+			func() {
+				defer func() {
+					if r := recover(); r != nil {
+						t.Logf("PeekPushNotificationName panicked as expected for insufficient data: %v", r)
+					}
+				}()
+				_, err := reader.PeekPushNotificationName()
+				if err == nil {
+					t.Error("PeekPushNotificationName should error for insufficient data")
+				}
+			}()
+		})
+
+		t.Run("EmptyBuffer", func(t *testing.T) {
+			buf := &bytes.Buffer{}
+			reader := NewReader(buf)
+
+			_, err := reader.PeekPushNotificationName()
+			if err == nil {
+				t.Error("PeekPushNotificationName should error for empty buffer")
+			}
+		})
+
+		t.Run("DifferentRESPTypes", func(t *testing.T) {
+			// Test with different RESP types that should be rejected
+			respTypes := []byte{'+', '-', ':', '$', '*', '%', '~', '|', '('}
+
+			for _, respType := range respTypes {
+				t.Run(fmt.Sprintf("Type_%c", respType), func(t *testing.T) {
+					buf := &bytes.Buffer{}
+					buf.WriteByte(respType)
+					buf.WriteString("test data that fills the buffer completely")
+					reader := NewReader(buf)
+
+					_, err := reader.PeekPushNotificationName()
+					if err == nil {
+						t.Errorf("PeekPushNotificationName should error for RESP type '%c'", respType)
+					}
+
+					// The error might be "no data available" or "can't parse push notification"
+					if !strings.Contains(err.Error(), "can't peek push notification name") {
+						t.Errorf("Error should mention push notification parsing, got: %v", err)
+					}
+				})
+			}
+		})
+	})
+
+	t.Run("EdgeCases", func(t *testing.T) {
+		t.Run("ZeroLengthArray", func(t *testing.T) {
+			// Create push notification with zero elements: >0\r\n
+			buf := &bytes.Buffer{}
+			buf.WriteString(">0\r\npadding_data_to_fill_buffer_completely")
+			reader := NewReader(buf)
+
+			_, err := reader.PeekPushNotificationName()
+			if err == nil {
+				t.Error("PeekPushNotificationName should error for zero-length array")
+			}
+		})
+
+		t.Run("EmptyNotificationName", func(t *testing.T) {
+			// Create push notification with empty name: >1\r\n$0\r\n\r\n
+			buf := createValidPushNotification("", "data")
+			reader := NewReader(buf)
+
+			// Prime the buffer
+			_, _ = reader.rd.Peek(1)
+
+			name, err := reader.PeekPushNotificationName()
+			if err != nil {
+				t.Errorf("PeekPushNotificationName should not error for empty name: %v", err)
+			}
+
+			if name != "" {
+				t.Errorf("Expected empty notification name, got '%s'", name)
+			}
+		})
+
+		t.Run("CorruptedData", func(t *testing.T) {
+			corruptedCases := []struct {
+				name string
+				data string
+			}{
+				{"CorruptedLength", ">abc\r\n$6\r\nMOVING\r\n"},
+				{"MissingCRLF", ">2$6\r\nMOVING\r\n$4\r\ndata\r\n"},
+				{"InvalidStringLength", ">2\r\n$abc\r\nMOVING\r\n$4\r\ndata\r\n"},
+				{"NegativeStringLength", ">2\r\n$-1\r\n$4\r\ndata\r\n"},
+				{"IncompleteString", ">1\r\n$6\r\nMOV"},
+			}
+
+			for _, tc := range corruptedCases {
+				t.Run(tc.name, func(t *testing.T) {
+					buf := &bytes.Buffer{}
+					buf.WriteString(tc.data)
+					reader := NewReader(buf)
+
+					// Some corrupted data might not error but return unexpected results
+					// This is acceptable behavior for malformed input
+					name, err := reader.PeekPushNotificationName()
+					if err != nil {
+						t.Logf("PeekPushNotificationName errored for corrupted data %s: %v (DATA: %s)", tc.name, err, tc.data)
+					} else {
+						t.Logf("PeekPushNotificationName returned '%s' for corrupted data NAME: %s, DATA: %s", name, tc.name, tc.data)
+					}
+				})
+			}
+		})
+	})
+
+	t.Run("BoundaryConditions", func(t *testing.T) {
+		t.Run("ExactlyPeekSize", func(t *testing.T) {
+			// Create buffer that is exactly 36 bytes (the peek window size)
+			buf := &bytes.Buffer{}
+			// ">1\r\n$4\r\nTEST\r\n" = 14 bytes, need 22 more
+			buf.WriteString(">1\r\n$4\r\nTEST\r\n1234567890123456789012")
+			if buf.Len() != 36 {
+				t.Errorf("Expected buffer length 36, got %d", buf.Len())
+			}
+
+			reader := NewReader(buf)
+			// Prime the buffer
+			_, _ = reader.rd.Peek(1)
+
+			name, err := reader.PeekPushNotificationName()
+			if err != nil {
+				t.Errorf("PeekPushNotificationName should work for exact peek size: %v", err)
+			}
+
+			if name != "TEST" {
+				t.Errorf("Expected 'TEST', got '%s'", name)
+			}
+		})
+
+		t.Run("LessThanPeekSize", func(t *testing.T) {
+			// Create buffer smaller than 36 bytes but with complete notification
+			buf := createValidPushNotification("TEST", "")
+			reader := NewReader(buf)
+
+			// Prime the buffer
+			_, _ = reader.rd.Peek(1)
+
+			name, err := reader.PeekPushNotificationName()
+			if err != nil {
+				t.Errorf("PeekPushNotificationName should work for complete notification: %v", err)
+			}
+
+			if name != "TEST" {
+				t.Errorf("Expected 'TEST', got '%s'", name)
+			}
+		})
+
+		t.Run("LongNotificationName", func(t *testing.T) {
+			// Test with notification name that might exceed peek window
+			longName := strings.Repeat("A", 20) // 20 character name (safe size)
+			buf := createValidPushNotification(longName, "data")
+			reader := NewReader(buf)
+
+			// Prime the buffer
+			_, _ = reader.rd.Peek(1)
+
+			name, err := reader.PeekPushNotificationName()
+			if err != nil {
+				t.Errorf("PeekPushNotificationName should work for long name: %v", err)
+			}
+
+			if name != longName {
+				t.Errorf("Expected '%s', got '%s'", longName, name)
+			}
+		})
+	})
+}
+
+// Helper functions to create test data
+
+// createValidPushNotification creates a valid RESP3 push notification
+func createValidPushNotification(notificationName, data string) *bytes.Buffer {
+	buf := &bytes.Buffer{}
+
+	simpleOrString := rand.Intn(2) == 0
+
+	if data == "" {
+
+		// Single element notification
+		buf.WriteString(">1\r\n")
+		if simpleOrString {
+			buf.WriteString(fmt.Sprintf("+%s\r\n", notificationName))
+		} else {
+			buf.WriteString(fmt.Sprintf("$%d\r\n%s\r\n", len(notificationName), notificationName))
+		}
+	} else {
+		// Two element notification
+		buf.WriteString(">2\r\n")
+		if simpleOrString {
+			buf.WriteString(fmt.Sprintf("+%s\r\n", notificationName))
+			buf.WriteString(fmt.Sprintf("+%s\r\n", data))
+		} else {
+			buf.WriteString(fmt.Sprintf("$%d\r\n%s\r\n", len(notificationName), notificationName))
+			buf.WriteString(fmt.Sprintf("$%d\r\n%s\r\n", len(notificationName), notificationName))
+		}
+	}
+
+	return buf
+}
+
+// createReaderWithPrimedBuffer creates a reader and primes the buffer
+func createReaderWithPrimedBuffer(buf *bytes.Buffer) *Reader {
+	reader := NewReader(buf)
+	// Prime the buffer by peeking first
+	_, _ = reader.rd.Peek(1)
+	return reader
+}
+
+// createPushNotificationWithArgs creates a push notification with multiple arguments
+func createPushNotificationWithArgs(notificationName string, args ...string) *bytes.Buffer {
+	buf := &bytes.Buffer{}
+
+	totalElements := 1 + len(args)
+	buf.WriteString(fmt.Sprintf(">%d\r\n", totalElements))
+
+	// Write notification name
+	buf.WriteString(fmt.Sprintf("$%d\r\n%s\r\n", len(notificationName), notificationName))
+
+	// Write arguments
+	for _, arg := range args {
+		buf.WriteString(fmt.Sprintf("$%d\r\n%s\r\n", len(arg), arg))
+	}
+
+	return buf
+}
+
+// createSingleElementPushNotification creates a push notification with single element
+func createSingleElementPushNotification(notificationName string) *bytes.Buffer {
+	buf := &bytes.Buffer{}
+	buf.WriteString(">1\r\n")
+	buf.WriteString(fmt.Sprintf("$%d\r\n%s\r\n", len(notificationName), notificationName))
+	return buf
+}
+
+// BenchmarkPeekPushNotificationName benchmarks the method performance
+func BenchmarkPeekPushNotificationName(b *testing.B) {
+	testCases := []struct {
+		name         string
+		notification string
+	}{
+		{"Short", "TEST"},
+		{"Medium", "MOVING_NOTIFICATION"},
+		{"Long", "VERY_LONG_NOTIFICATION_NAME_FOR_TESTING"},
+	}
+
+	for _, tc := range testCases {
+		b.Run(tc.name, func(b *testing.B) {
+			buf := createValidPushNotification(tc.notification, "data")
+			data := buf.Bytes()
+
+			b.ResetTimer()
+			for i := 0; i < b.N; i++ {
+				reader := NewReader(bytes.NewReader(data))
+				_, err := reader.PeekPushNotificationName()
+				if err != nil {
+					b.Errorf("PeekPushNotificationName should not error: %v", err)
+				}
+			}
+		})
+	}
+}
+
+// TestPeekPushNotificationNameSpecialCases tests special cases and realistic scenarios
+func TestPeekPushNotificationNameSpecialCases(t *testing.T) {
+	t.Run("RealisticNotifications", func(t *testing.T) {
+		// Test realistic Redis push notifications
+		realisticCases := []struct {
+			name         string
+			notification []string
+			expected     string
+		}{
+			{"MovingSlot", []string{"MOVING", "slot", "123", "from", "127.0.0.1:7000", "to", "127.0.0.1:7001"}, "MOVING"},
+			{"MigratingSlot", []string{"MIGRATING", "slot", "456", "from", "127.0.0.1:7001", "to", "127.0.0.1:7002"}, "MIGRATING"},
+			{"MigratedSlot", []string{"MIGRATED", "slot", "789", "from", "127.0.0.1:7002", "to", "127.0.0.1:7000"}, "MIGRATED"},
+			{"FailingOver", []string{"FAILING_OVER", "node", "127.0.0.1:7000"}, "FAILING_OVER"},
+			{"FailedOver", []string{"FAILED_OVER", "node", "127.0.0.1:7000"}, "FAILED_OVER"},
+			{"PubSubMessage", []string{"message", "mychannel", "hello world"}, "message"},
+			{"PubSubPMessage", []string{"pmessage", "pattern*", "mychannel", "hello world"}, "pmessage"},
+			{"Subscribe", []string{"subscribe", "mychannel", "1"}, "subscribe"},
+			{"Unsubscribe", []string{"unsubscribe", "mychannel", "0"}, "unsubscribe"},
+		}
+
+		for _, tc := range realisticCases {
+			t.Run(tc.name, func(t *testing.T) {
+				buf := createPushNotificationWithArgs(tc.notification[0], tc.notification[1:]...)
+				reader := createReaderWithPrimedBuffer(buf)
+
+				name, err := reader.PeekPushNotificationName()
+				if err != nil {
+					t.Errorf("PeekPushNotificationName should not error for %s: %v", tc.name, err)
+				}
+
+				if name != tc.expected {
+					t.Errorf("Expected '%s', got '%s'", tc.expected, name)
+				}
+			})
+		}
+	})
+
+	t.Run("SpecialCharactersInName", func(t *testing.T) {
+		specialCases := []struct {
+			name         string
+			notification string
+		}{
+			{"WithUnderscore", "test_notification"},
+			{"WithDash", "test-notification"},
+			{"WithNumbers", "test123"},
+			{"WithDots", "test.notification"},
+			{"WithColon", "test:notification"},
+			{"WithSlash", "test/notification"},
+			{"MixedCase", "TestNotification"},
+			{"AllCaps", "TESTNOTIFICATION"},
+			{"AllLower", "testnotification"},
+			{"Unicode", "tëst"},
+		}
+
+		for _, tc := range specialCases {
+			t.Run(tc.name, func(t *testing.T) {
+				buf := createValidPushNotification(tc.notification, "data")
+				reader := createReaderWithPrimedBuffer(buf)
+
+				name, err := reader.PeekPushNotificationName()
+				if err != nil {
+					t.Errorf("PeekPushNotificationName should not error for '%s': %v", tc.notification, err)
+				}
+
+				if name != tc.notification {
+					t.Errorf("Expected '%s', got '%s'", tc.notification, name)
+				}
+			})
+		}
+	})
+
+	t.Run("IdempotentPeek", func(t *testing.T) {
+		// Test that multiple peeks return the same result
+		buf := createValidPushNotification("MOVING", "data")
+		reader := createReaderWithPrimedBuffer(buf)
+
+		// First peek
+		name1, err1 := reader.PeekPushNotificationName()
+		if err1 != nil {
+			t.Errorf("First PeekPushNotificationName should not error: %v", err1)
+		}
+
+		// Second peek should return the same result
+		name2, err2 := reader.PeekPushNotificationName()
+		if err2 != nil {
+			t.Errorf("Second PeekPushNotificationName should not error: %v", err2)
+		}
+
+		if name1 != name2 {
+			t.Errorf("Peek should be idempotent: first='%s', second='%s'", name1, name2)
+		}
+
+		if name1 != "MOVING" {
+			t.Errorf("Expected 'MOVING', got '%s'", name1)
+		}
+	})
+}
+
+// TestPeekPushNotificationNamePerformance tests performance characteristics
+func TestPeekPushNotificationNamePerformance(t *testing.T) {
+	t.Run("RepeatedCalls", func(t *testing.T) {
+		// Test that repeated calls work correctly
+		buf := createValidPushNotification("TEST", "data")
+		reader := createReaderWithPrimedBuffer(buf)
+
+		// Call multiple times
+		for i := 0; i < 10; i++ {
+			name, err := reader.PeekPushNotificationName()
+			if err != nil {
+				t.Errorf("PeekPushNotificationName should not error on call %d: %v", i, err)
+			}
+			if name != "TEST" {
+				t.Errorf("Expected 'TEST' on call %d, got '%s'", i, name)
+			}
+		}
+	})
+
+	t.Run("LargeNotifications", func(t *testing.T) {
+		// Test with large notification data
+		largeData := strings.Repeat("x", 1000)
+		buf := createValidPushNotification("LARGE", largeData)
+		reader := createReaderWithPrimedBuffer(buf)
+
+		name, err := reader.PeekPushNotificationName()
+		if err != nil {
+			t.Errorf("PeekPushNotificationName should not error for large notification: %v", err)
+		}
+
+		if name != "LARGE" {
+			t.Errorf("Expected 'LARGE', got '%s'", name)
+		}
+	})
+}
+
+// TestPeekPushNotificationNameBehavior documents the method's behavior
+func TestPeekPushNotificationNameBehavior(t *testing.T) {
+	t.Run("MethodBehavior", func(t *testing.T) {
+		// Test that the method works as intended:
+		// 1. Peek at the buffer without consuming it
+		// 2. Detect push notifications (RESP type '>')
+		// 3. Extract the notification name from the first element
+		// 4. Return the name for filtering decisions
+
+		buf := createValidPushNotification("MOVING", "slot_data")
+		reader := createReaderWithPrimedBuffer(buf)
+
+		// Peek should not consume the buffer
+		name, err := reader.PeekPushNotificationName()
+		if err != nil {
+			t.Errorf("PeekPushNotificationName should not error: %v", err)
+		}
+
+		if name != "MOVING" {
+			t.Errorf("Expected 'MOVING', got '%s'", name)
+		}
+
+		// Buffer should still be available for normal reading
+		replyType, err := reader.PeekReplyType()
+		if err != nil {
+			t.Errorf("PeekReplyType should work after PeekPushNotificationName: %v", err)
+		}
+
+		if replyType != RespPush {
+			t.Errorf("Expected RespPush, got %v", replyType)
+		}
+	})
+
+	t.Run("BufferNotConsumed", func(t *testing.T) {
+		// Verify that peeking doesn't consume the buffer
+		buf := createValidPushNotification("TEST", "data")
+		originalData := buf.Bytes()
+		reader := createReaderWithPrimedBuffer(buf)
+
+		// Peek the notification name
+		name, err := reader.PeekPushNotificationName()
+		if err != nil {
+			t.Errorf("PeekPushNotificationName should not error: %v", err)
+		}
+
+		if name != "TEST" {
+			t.Errorf("Expected 'TEST', got '%s'", name)
+		}
+
+		// Read the actual notification
+		reply, err := reader.ReadReply()
+		if err != nil {
+			t.Errorf("ReadReply should work after peek: %v", err)
+		}
+
+		// Verify we got the complete notification
+		if replySlice, ok := reply.([]interface{}); ok {
+			if len(replySlice) != 2 {
+				t.Errorf("Expected 2 elements, got %d", len(replySlice))
+			}
+			if replySlice[0] != "TEST" {
+				t.Errorf("Expected 'TEST', got %v", replySlice[0])
+			}
+		} else {
+			t.Errorf("Expected slice reply, got %T", reply)
+		}
+
+		// Verify buffer was properly consumed
+		if buf.Len() != 0 {
+			t.Errorf("Buffer should be empty after reading, but has %d bytes: %q", buf.Len(), buf.Bytes())
+		}
+
+		t.Logf("Original buffer size: %d bytes", len(originalData))
+		t.Logf("Successfully peeked and then read complete notification")
+	})
+
+	t.Run("ImplementationSuccess", func(t *testing.T) {
+		// Document that the implementation is now working correctly
+		t.Log("PeekPushNotificationName implementation status:")
+		t.Log("1. ✅ Correctly parses RESP3 push notifications")
+		t.Log("2. ✅ Extracts notification names properly")
+		t.Log("3. ✅ Handles buffer peeking without consumption")
+		t.Log("4. ✅ Works with various notification types")
+		t.Log("5. ✅ Supports empty notification names")
+		t.Log("")
+		t.Log("RESP3 format parsing:")
+		t.Log(">2\\r\\n$6\\r\\nMOVING\\r\\n$4\\r\\ndata\\r\\n")
+		t.Log("✅ Correctly identifies push notification marker (>)")
+		t.Log("✅ Skips array length (2)")
+		t.Log("✅ Parses string marker ($) and length (6)")
+		t.Log("✅ Extracts notification name (MOVING)")
+		t.Log("✅ Returns name without consuming buffer")
+		t.Log("")
+		t.Log("Note: Buffer must be primed with a peek operation first")
+	})
+}

internal/proto/reader.go

@@ -99,6 +99,92 @@ func (r *Reader) PeekReplyType() (byte, error) {
 	return b[0], nil
 }
 
+func (r *Reader) PeekPushNotificationName() (string, error) {
+	// "prime" the buffer by peeking at the next byte
+	c, err := r.Peek(1)
+	if err != nil {
+		return "", err
+	}
+	if c[0] != RespPush {
+		return "", fmt.Errorf("redis: can't peek push notification name, next reply is not a push notification")
+	}
+
+	// peek 36 bytes at most, should be enough to read the push notification name
+	toPeek := 36
+	buffered := r.Buffered()
+	if buffered == 0 {
+		return "", fmt.Errorf("redis: can't peek push notification name, no data available")
+	}
+	if buffered < toPeek {
+		toPeek = buffered
+	}
+	buf, err := r.rd.Peek(toPeek)
+	if err != nil {
+		return "", err
+	}
+	if buf[0] != RespPush {
+		return "", fmt.Errorf("redis: can't parse push notification: %q", buf)
+	}
+
+	if len(buf) < 3 {
+		return "", fmt.Errorf("redis: can't parse push notification: %q", buf)
+	}
+
+	// remove push notification type
+	buf = buf[1:]
+	// remove first line - e.g. >2\r\n
+	for i := 0; i < len(buf)-1; i++ {
+		if buf[i] == '\r' && buf[i+1] == '\n' {
+			buf = buf[i+2:]
+			break
+		} else {
+			if buf[i] < '0' || buf[i] > '9' {
+				return "", fmt.Errorf("redis: can't parse push notification: %q", buf)
+			}
+		}
+	}
+	if len(buf) < 2 {
+		return "", fmt.Errorf("redis: can't parse push notification: %q", buf)
+	}
+	// next line should be $<length><string>\r\n or +<length><string>\r\n
+	// should have the type of the push notification name and it's length
+	if buf[0] != RespString && buf[0] != RespStatus {
+		return "", fmt.Errorf("redis: can't parse push notification name: %q", buf)
+	}
+	typeOfName := buf[0]
+	// remove the type of the push notification name
+	buf = buf[1:]
+	if typeOfName == RespString {
+		// remove the length of the string
+		if len(buf) < 2 {
+			return "", fmt.Errorf("redis: can't parse push notification name: %q", buf)
+		}
+		for i := 0; i < len(buf)-1; i++ {
+			if buf[i] == '\r' && buf[i+1] == '\n' {
+				buf = buf[i+2:]
+				break
+			} else {
+				if buf[i] < '0' || buf[i] > '9' {
+					return "", fmt.Errorf("redis: can't parse push notification name: %q", buf)
+				}
+			}
+		}
+	}
+
+	if len(buf) < 2 {
+		return "", fmt.Errorf("redis: can't parse push notification name: %q", buf)
+	}
+	// keep only the notification name
+	for i := 0; i < len(buf)-1; i++ {
+		if buf[i] == '\r' && buf[i+1] == '\n' {
+			buf = buf[:i]
+			break
+		}
+	}
+
+	return util.BytesToString(buf), nil
+}
+
 // ReadLine Return a valid reply, it will check the protocol or redis error,
 // and discard the attribute type.
 func (r *Reader) ReadLine() ([]byte, error) {

internal/redis.go

@@ -0,0 +1,3 @@
+package internal
+
+const RedisNull = "null"

internal/util/convert.go

@@ -28,3 +28,14 @@ func MustParseFloat(s string) float64 {
 	}
 	return f
 }
+
+// SafeIntToInt32 safely converts an int to int32, returning an error if overflow would occur.
+func SafeIntToInt32(value int, fieldName string) (int32, error) {
+	if value > math.MaxInt32 {
+		return 0, fmt.Errorf("redis: %s value %d exceeds maximum allowed value %d", fieldName, value, math.MaxInt32)
+	}
+	if value < math.MinInt32 {
+		return 0, fmt.Errorf("redis: %s value %d is below minimum allowed value %d", fieldName, value, math.MinInt32)
+	}
+	return int32(value), nil
+}

internal/util/math.go

@@ -0,0 +1,17 @@
+package util
+
+// Max returns the maximum of two integers
+func Max(a, b int) int {
+	if a > b {
+		return a
+	}
+	return b
+}
+
+// Min returns the minimum of two integers
+func Min(a, b int) int {
+	if a < b {
+		return a
+	}
+	return b
+}

internal_test.go

@@ -16,6 +16,8 @@ import (
 	. "github.com/bsm/gomega"
 )
 
+var ctx = context.TODO()
+
 var _ = Describe("newClusterState", func() {
 	var state *clusterState
 

logging/logging.go

@@ -0,0 +1,121 @@
+// Package logging provides logging level constants and utilities for the go-redis library.
+// This package centralizes logging configuration to ensure consistency across all components.
+package logging
+
+import (
+	"context"
+	"fmt"
+	"strings"
+
+	"github.com/redis/go-redis/v9/internal"
+)
+
+// LogLevel represents the logging level
+type LogLevel int
+
+// Log level constants for the entire go-redis library
+const (
+	LogLevelError LogLevel = iota // 0 - errors only
+	LogLevelWarn                  // 1 - warnings and errors
+	LogLevelInfo                  // 2 - info, warnings, and errors
+	LogLevelDebug                 // 3 - debug, info, warnings, and errors
+)
+
+// String returns the string representation of the log level
+func (l LogLevel) String() string {
+	switch l {
+	case LogLevelError:
+		return "ERROR"
+	case LogLevelWarn:
+		return "WARN"
+	case LogLevelInfo:
+		return "INFO"
+	case LogLevelDebug:
+		return "DEBUG"
+	default:
+		return "UNKNOWN"
+	}
+}
+
+// IsValid returns true if the log level is valid
+func (l LogLevel) IsValid() bool {
+	return l >= LogLevelError && l <= LogLevelDebug
+}
+
+func (l LogLevel) WarnOrAbove() bool {
+	return l >= LogLevelWarn
+}
+
+func (l LogLevel) InfoOrAbove() bool {
+	return l >= LogLevelInfo
+}
+
+func (l LogLevel) DebugOrAbove() bool {
+	return l >= LogLevelDebug
+}
+
+// VoidLogger is a logger that does nothing.
+// Used to disable logging and thus speed up the library.
+type VoidLogger struct{}
+
+func (v *VoidLogger) Printf(_ context.Context, _ string, _ ...interface{}) {
+	// do nothing
+}
+
+// Disable disables logging by setting the internal logger to a void logger.
+// This can be used to speed up the library if logging is not needed.
+// It will override any custom logger that was set before and set the VoidLogger.
+func Disable() {
+	internal.Logger = &VoidLogger{}
+}
+
+// Enable enables logging by setting the internal logger to the default logger.
+// This is the default behavior.
+// You can use redis.SetLogger to set a custom logger.
+//
+// NOTE: This function is not thread-safe.
+// It will override any custom logger that was set before and set the DefaultLogger.
+func Enable() {
+	internal.Logger = internal.NewDefaultLogger()
+}
+
+// NewBlacklistLogger returns a new logger that filters out messages containing any of the substrings.
+// This can be used to filter out messages containing sensitive information.
+func NewBlacklistLogger(substr []string) internal.Logging {
+	l := internal.NewDefaultLogger()
+	return &filterLogger{logger: l, substr: substr, blacklist: true}
+}
+
+// NewWhitelistLogger returns a new logger that only logs messages containing any of the substrings.
+// This can be used to only log messages related to specific commands or patterns.
+func NewWhitelistLogger(substr []string) internal.Logging {
+	l := internal.NewDefaultLogger()
+	return &filterLogger{logger: l, substr: substr, blacklist: false}
+}
+
+type filterLogger struct {
+	logger    internal.Logging
+	blacklist bool
+	substr    []string
+}
+
+func (l *filterLogger) Printf(ctx context.Context, format string, v ...interface{}) {
+	msg := fmt.Sprintf(format, v...)
+	found := false
+	for _, substr := range l.substr {
+		if strings.Contains(msg, substr) {
+			found = true
+			if l.blacklist {
+				return
+			}
+		}
+	}
+	// whitelist, only log if one of the substrings is present
+	if !l.blacklist && !found {
+		return
+	}
+	if l.logger != nil {
+		l.logger.Printf(ctx, format, v...)
+		return
+	}
+}

logging/logging_test.go

@@ -0,0 +1,59 @@
+package logging
+
+import "testing"
+
+func TestLogLevel_String(t *testing.T) {
+	tests := []struct {
+		level    LogLevel
+		expected string
+	}{
+		{LogLevelError, "ERROR"},
+		{LogLevelWarn, "WARN"},
+		{LogLevelInfo, "INFO"},
+		{LogLevelDebug, "DEBUG"},
+		{LogLevel(99), "UNKNOWN"},
+	}
+
+	for _, test := range tests {
+		if got := test.level.String(); got != test.expected {
+			t.Errorf("LogLevel(%d).String() = %q, want %q", test.level, got, test.expected)
+		}
+	}
+}
+
+func TestLogLevel_IsValid(t *testing.T) {
+	tests := []struct {
+		level    LogLevel
+		expected bool
+	}{
+		{LogLevelError, true},
+		{LogLevelWarn, true},
+		{LogLevelInfo, true},
+		{LogLevelDebug, true},
+		{LogLevel(-1), false},
+		{LogLevel(4), false},
+		{LogLevel(99), false},
+	}
+
+	for _, test := range tests {
+		if got := test.level.IsValid(); got != test.expected {
+			t.Errorf("LogLevel(%d).IsValid() = %v, want %v", test.level, got, test.expected)
+		}
+	}
+}
+
+func TestLogLevelConstants(t *testing.T) {
+	// Test that constants have expected values
+	if LogLevelError != 0 {
+		t.Errorf("LogLevelError = %d, want 0", LogLevelError)
+	}
+	if LogLevelWarn != 1 {
+		t.Errorf("LogLevelWarn = %d, want 1", LogLevelWarn)
+	}
+	if LogLevelInfo != 2 {
+		t.Errorf("LogLevelInfo = %d, want 2", LogLevelInfo)
+	}
+	if LogLevelDebug != 3 {
+		t.Errorf("LogLevelDebug = %d, want 3", LogLevelDebug)
+	}
+}

main_test.go

@@ -13,6 +13,7 @@ import (
 	. "github.com/bsm/ginkgo/v2"
 	. "github.com/bsm/gomega"
 	"github.com/redis/go-redis/v9"
+	"github.com/redis/go-redis/v9/logging"
 )
 
 const (
@@ -102,6 +103,7 @@ var _ = BeforeSuite(func() {
 	fmt.Printf("RCEDocker: %v\n", RCEDocker)
 	fmt.Printf("REDIS_VERSION: %.1f\n", RedisVersion)
 	fmt.Printf("CLIENT_LIBS_TEST_IMAGE: %v\n", os.Getenv("CLIENT_LIBS_TEST_IMAGE"))
+	logging.Disable()
 
 	if RedisVersion < 7.0 || RedisVersion > 9 {
 		panic("incorrect or not supported redis version")

options.go

@@ -14,8 +14,11 @@ import (
 	"time"
 
 	"github.com/redis/go-redis/v9/auth"
+	"github.com/redis/go-redis/v9/hitless"
 	"github.com/redis/go-redis/v9/internal/pool"
 	"github.com/redis/go-redis/v9/internal/proto"
+	"github.com/redis/go-redis/v9/internal/util"
+	"github.com/redis/go-redis/v9/push"
 )
 
 // Limiter is the interface of a rate limiter or a circuit breaker.
@@ -109,6 +112,16 @@ type Options struct {
 	// default: 5 seconds
 	DialTimeout time.Duration
 
+	// DialerRetries is the maximum number of retry attempts when dialing fails.
+	//
+	// default: 5
+	DialerRetries int
+
+	// DialerRetryTimeout is the backoff duration between retry attempts.
+	//
+	// default: 100 milliseconds
+	DialerRetryTimeout time.Duration
+
 	// ReadTimeout for socket reads. If reached, commands will fail
 	// with a timeout instead of blocking. Supported values:
 	//
@@ -152,6 +165,7 @@ type Options struct {
 	//
 	// Note that FIFO has slightly higher overhead compared to LIFO,
 	// but it helps closing idle connections faster reducing the pool size.
+	// default: false
 	PoolFIFO bool
 
 	// PoolSize is the base number of socket connections.
@@ -232,12 +246,30 @@ type Options struct {
 	// When unstable mode is enabled, the client will use RESP3 protocol and only be able to use RawResult
 	UnstableResp3 bool
 
+	// Push notifications are always enabled for RESP3 connections (Protocol: 3)
+	// and are not available for RESP2 connections. No configuration option is needed.
+
+	// PushNotificationProcessor is the processor for handling push notifications.
+	// If nil, a default processor will be created for RESP3 connections.
+	PushNotificationProcessor push.NotificationProcessor
+
 	// FailingTimeoutSeconds is the timeout in seconds for marking a cluster node as failing.
 	// When a node is marked as failing, it will be avoided for this duration.
 	// Default is 15 seconds.
 	FailingTimeoutSeconds int
+
+	// HitlessUpgradeConfig provides custom configuration for hitless upgrades.
+	// When HitlessUpgradeConfig.Mode is not "disabled", the client will handle
+	// cluster upgrade notifications gracefully and manage connection/pool state
+	// transitions seamlessly. Requires Protocol: 3 (RESP3) for push notifications.
+	// If nil, hitless upgrades are in "auto" mode and will be enabled if the server supports it.
+	HitlessUpgradeConfig *HitlessUpgradeConfig
 }
 
+// HitlessUpgradeConfig provides configuration options for hitless upgrades.
+// This is an alias to hitless.Config for convenience.
+type HitlessUpgradeConfig = hitless.Config
+
 func (opt *Options) init() {
 	if opt.Addr == "" {
 		opt.Addr = "localhost:6379"
@@ -255,6 +287,12 @@ func (opt *Options) init() {
 	if opt.DialTimeout == 0 {
 		opt.DialTimeout = 5 * time.Second
 	}
+	if opt.DialerRetries == 0 {
+		opt.DialerRetries = 5
+	}
+	if opt.DialerRetryTimeout == 0 {
+		opt.DialerRetryTimeout = 100 * time.Millisecond
+	}
 	if opt.Dialer == nil {
 		opt.Dialer = NewDialer(opt)
 	}
@@ -312,13 +350,36 @@ func (opt *Options) init() {
 	case 0:
 		opt.MaxRetryBackoff = 512 * time.Millisecond
 	}
+
+	opt.HitlessUpgradeConfig = opt.HitlessUpgradeConfig.ApplyDefaultsWithPoolConfig(opt.PoolSize, opt.MaxActiveConns)
+
+	// auto-detect endpoint type if not specified
+	endpointType := opt.HitlessUpgradeConfig.EndpointType
+	if endpointType == "" || endpointType == hitless.EndpointTypeAuto {
+		// Auto-detect endpoint type if not specified
+		endpointType = hitless.DetectEndpointType(opt.Addr, opt.TLSConfig != nil)
+	}
+	opt.HitlessUpgradeConfig.EndpointType = endpointType
 }
 
 func (opt *Options) clone() *Options {
 	clone := *opt
+
+	// Deep clone HitlessUpgradeConfig to avoid sharing between clients
+	if opt.HitlessUpgradeConfig != nil {
+		configClone := *opt.HitlessUpgradeConfig
+		clone.HitlessUpgradeConfig = &configClone
+	}
+
 	return &clone
 }
 
+// NewDialer returns a function that will be used as the default dialer
+// when none is specified in Options.Dialer.
+func (opt *Options) NewDialer() func(context.Context, string, string) (net.Conn, error) {
+	return NewDialer(opt)
+}
+
 // NewDialer returns a function that will be used as the default dialer
 // when none is specified in Options.Dialer.
 func NewDialer(opt *Options) func(context.Context, string, string) (net.Conn, error) {
@@ -604,21 +665,84 @@ func getUserPassword(u *url.URL) (string, string) {
 func newConnPool(
 	opt *Options,
 	dialer func(ctx context.Context, network, addr string) (net.Conn, error),
-) *pool.ConnPool {
+) (*pool.ConnPool, error) {
+	poolSize, err := util.SafeIntToInt32(opt.PoolSize, "PoolSize")
+	if err != nil {
+		return nil, err
+	}
+
+	minIdleConns, err := util.SafeIntToInt32(opt.MinIdleConns, "MinIdleConns")
+	if err != nil {
+		return nil, err
+	}
+
+	maxIdleConns, err := util.SafeIntToInt32(opt.MaxIdleConns, "MaxIdleConns")
+	if err != nil {
+		return nil, err
+	}
+
+	maxActiveConns, err := util.SafeIntToInt32(opt.MaxActiveConns, "MaxActiveConns")
+	if err != nil {
+		return nil, err
+	}
+
 	return pool.NewConnPool(&pool.Options{
 		Dialer: func(ctx context.Context) (net.Conn, error) {
 			return dialer(ctx, opt.Network, opt.Addr)
 		},
 		PoolFIFO:                 opt.PoolFIFO,
-		PoolSize:        opt.PoolSize,
+		PoolSize:                 poolSize,
 		PoolTimeout:              opt.PoolTimeout,
 		DialTimeout:              opt.DialTimeout,
-		MinIdleConns:    opt.MinIdleConns,
-		MaxIdleConns:    opt.MaxIdleConns,
-		MaxActiveConns:  opt.MaxActiveConns,
+		DialerRetries:            opt.DialerRetries,
+		DialerRetryTimeout:       opt.DialerRetryTimeout,
+		MinIdleConns:             minIdleConns,
+		MaxIdleConns:             maxIdleConns,
+		MaxActiveConns:           maxActiveConns,
 		ConnMaxIdleTime:          opt.ConnMaxIdleTime,
 		ConnMaxLifetime:          opt.ConnMaxLifetime,
 		ReadBufferSize:           opt.ReadBufferSize,
 		WriteBufferSize:          opt.WriteBufferSize,
-	})
+		PushNotificationsEnabled: opt.Protocol == 3,
+	}), nil
+}
+
+func newPubSubPool(opt *Options, dialer func(ctx context.Context, network, addr string) (net.Conn, error),
+) (*pool.PubSubPool, error) {
+	poolSize, err := util.SafeIntToInt32(opt.PoolSize, "PoolSize")
+	if err != nil {
+		return nil, err
+	}
+
+	minIdleConns, err := util.SafeIntToInt32(opt.MinIdleConns, "MinIdleConns")
+	if err != nil {
+		return nil, err
+	}
+
+	maxIdleConns, err := util.SafeIntToInt32(opt.MaxIdleConns, "MaxIdleConns")
+	if err != nil {
+		return nil, err
+	}
+
+	maxActiveConns, err := util.SafeIntToInt32(opt.MaxActiveConns, "MaxActiveConns")
+	if err != nil {
+		return nil, err
+	}
+
+	return pool.NewPubSubPool(&pool.Options{
+		PoolFIFO:                 opt.PoolFIFO,
+		PoolSize:                 poolSize,
+		PoolTimeout:              opt.PoolTimeout,
+		DialTimeout:              opt.DialTimeout,
+		DialerRetries:            opt.DialerRetries,
+		DialerRetryTimeout:       opt.DialerRetryTimeout,
+		MinIdleConns:             minIdleConns,
+		MaxIdleConns:             maxIdleConns,
+		MaxActiveConns:           maxActiveConns,
+		ConnMaxIdleTime:          opt.ConnMaxIdleTime,
+		ConnMaxLifetime:          opt.ConnMaxLifetime,
+		ReadBufferSize:           32 * 1024,
+		WriteBufferSize:          32 * 1024,
+		PushNotificationsEnabled: opt.Protocol == 3,
+	}, dialer), nil
 }

osscluster.go

@@ -20,6 +20,7 @@ import (
 	"github.com/redis/go-redis/v9/internal/pool"
 	"github.com/redis/go-redis/v9/internal/proto"
 	"github.com/redis/go-redis/v9/internal/rand"
+	"github.com/redis/go-redis/v9/push"
 )
 
 const (
@@ -38,6 +39,7 @@ type ClusterOptions struct {
 	ClientName string
 
 	// NewClient creates a cluster node client with provided name and options.
+	// If NewClient is set by the user, the user is responsible for handling hitless upgrades and push notifications.
 	NewClient func(opt *Options) *Client
 
 	// The maximum number of retries before giving up. Command is retried
@@ -125,10 +127,22 @@ type ClusterOptions struct {
 	// UnstableResp3 enables Unstable mode for Redis Search module with RESP3.
 	UnstableResp3 bool
 
+	// PushNotificationProcessor is the processor for handling push notifications.
+	// If nil, a default processor will be created for RESP3 connections.
+	PushNotificationProcessor push.NotificationProcessor
+
 	// FailingTimeoutSeconds is the timeout in seconds for marking a cluster node as failing.
 	// When a node is marked as failing, it will be avoided for this duration.
 	// Default is 15 seconds.
 	FailingTimeoutSeconds int
+
+	// HitlessUpgradeConfig provides custom configuration for hitless upgrades.
+	// When HitlessUpgradeConfig.Mode is not "disabled", the client will handle
+	// cluster upgrade notifications gracefully and manage connection/pool state
+	// transitions seamlessly. Requires Protocol: 3 (RESP3) for push notifications.
+	// If nil, hitless upgrades are in "auto" mode and will be enabled if the server supports it.
+	// The ClusterClient does not directly work with hitless, it is up to the clients in the Nodes map to work with hitless.
+	HitlessUpgradeConfig *HitlessUpgradeConfig
 }
 
 func (opt *ClusterOptions) init() {
@@ -319,6 +333,13 @@ func setupClusterQueryParams(u *url.URL, o *ClusterOptions) (*ClusterOptions, er
 }
 
 func (opt *ClusterOptions) clientOptions() *Options {
+	// Clone HitlessUpgradeConfig to avoid sharing between cluster node clients
+	var hitlessConfig *HitlessUpgradeConfig
+	if opt.HitlessUpgradeConfig != nil {
+		configClone := *opt.HitlessUpgradeConfig
+		hitlessConfig = &configClone
+	}
+
 	return &Options{
 		ClientName: opt.ClientName,
 		Dialer:     opt.Dialer,
@@ -362,6 +383,8 @@ func (opt *ClusterOptions) clientOptions() *Options {
 		// situations in the options below will prevent that from happening.
 		readOnly:                  opt.ReadOnly && opt.ClusterSlots == nil,
 		UnstableResp3:             opt.UnstableResp3,
+		HitlessUpgradeConfig:      hitlessConfig,
+		PushNotificationProcessor: opt.PushNotificationProcessor,
 	}
 }
 
@@ -1664,7 +1687,7 @@ func (c *ClusterClient) processTxPipelineNode(
 }
 
 func (c *ClusterClient) processTxPipelineNodeConn(
-	ctx context.Context, _ *clusterNode, cn *pool.Conn, cmds []Cmder, failedCmds *cmdsMap,
+	ctx context.Context, node *clusterNode, cn *pool.Conn, cmds []Cmder, failedCmds *cmdsMap,
 ) error {
 	if err := cn.WithWriter(c.context(ctx), c.opt.WriteTimeout, func(wr *proto.Writer) error {
 		return writeCmds(wr, cmds)
@@ -1682,7 +1705,7 @@ func (c *ClusterClient) processTxPipelineNodeConn(
 		trimmedCmds := cmds[1 : len(cmds)-1]
 
 		if err := c.txPipelineReadQueued(
-			ctx, rd, statusCmd, trimmedCmds, failedCmds,
+			ctx, node, cn, rd, statusCmd, trimmedCmds, failedCmds,
 		); err != nil {
 			setCmdsErr(cmds, err)
 
@@ -1694,23 +1717,37 @@ func (c *ClusterClient) processTxPipelineNodeConn(
 			return err
 		}
 
-		return pipelineReadCmds(rd, trimmedCmds)
+		return node.Client.pipelineReadCmds(ctx, cn, rd, trimmedCmds)
 	})
 }
 
 func (c *ClusterClient) txPipelineReadQueued(
 	ctx context.Context,
+	node *clusterNode,
+	cn *pool.Conn,
 	rd *proto.Reader,
 	statusCmd *StatusCmd,
 	cmds []Cmder,
 	failedCmds *cmdsMap,
 ) error {
 	// Parse queued replies.
+	// To be sure there are no buffered push notifications, we process them before reading the reply
+	if err := node.Client.processPendingPushNotificationWithReader(ctx, cn, rd); err != nil {
+		// Log the error but don't fail the command execution
+		// Push notification processing errors shouldn't break normal Redis operations
+		internal.Logger.Printf(ctx, "push: error processing pending notifications before reading reply: %v", err)
+	}
 	if err := statusCmd.readReply(rd); err != nil {
 		return err
 	}
 
 	for _, cmd := range cmds {
+		// To be sure there are no buffered push notifications, we process them before reading the reply
+		if err := node.Client.processPendingPushNotificationWithReader(ctx, cn, rd); err != nil {
+			// Log the error but don't fail the command execution
+			// Push notification processing errors shouldn't break normal Redis operations
+			internal.Logger.Printf(ctx, "push: error processing pending notifications before reading reply: %v", err)
+		}
 		err := statusCmd.readReply(rd)
 		if err != nil {
 			if c.checkMovedErr(ctx, cmd, err, failedCmds) {
@@ -1724,6 +1761,12 @@ func (c *ClusterClient) txPipelineReadQueued(
 		}
 	}
 
+	// To be sure there are no buffered push notifications, we process them before reading the reply
+	if err := node.Client.processPendingPushNotificationWithReader(ctx, cn, rd); err != nil {
+		// Log the error but don't fail the command execution
+		// Push notification processing errors shouldn't break normal Redis operations
+		internal.Logger.Printf(ctx, "push: error processing pending notifications before reading reply: %v", err)
+	}
 	// Parse number of replies.
 	line, err := rd.ReadLine()
 	if err != nil {
@@ -1829,12 +1872,12 @@ func (c *ClusterClient) Watch(ctx context.Context, fn func(*Tx) error, keys ...s
 	return err
 }
 
+// hitless won't work here for now
 func (c *ClusterClient) pubSub() *PubSub {
 	var node *clusterNode
 	pubsub := &PubSub{
 		opt: c.opt.clientOptions(),
-
-		newConn: func(ctx context.Context, channels []string) (*pool.Conn, error) {
+		newConn: func(ctx context.Context, addr string, channels []string) (*pool.Conn, error) {
 			if node != nil {
 				panic("node != nil")
 			}
@@ -1868,18 +1911,25 @@ func (c *ClusterClient) pubSub() *PubSub {
 					return nil, err
 				}
 			}
-
-			cn, err := node.Client.newConn(context.TODO())
+			cn, err := node.Client.pubSubPool.NewConn(ctx, node.Client.opt.Network, node.Client.opt.Addr, channels)
 			if err != nil {
 				node = nil
-
 				return nil, err
 			}
-
+			// will return nil if already initialized
+			err = node.Client.initConn(ctx, cn)
+			if err != nil {
+				_ = cn.Close()
+				node = nil
+				return nil, err
+			}
+			node.Client.pubSubPool.TrackConn(cn)
 			return cn, nil
 		},
 		closeConn: func(cn *pool.Conn) error {
-			err := node.Client.connPool.CloseConn(cn)
+			// Untrack connection from PubSubPool
+			node.Client.pubSubPool.UntrackConn(cn)
+			err := cn.Close()
 			node = nil
 			return err
 		},

pool_pubsub_bench_test.go

@@ -0,0 +1,375 @@
+// Pool and PubSub Benchmark Suite
+//
+// This file contains comprehensive benchmarks for both pool operations and PubSub initialization.
+// It's designed to be run against different branches to compare performance.
+//
+// Usage Examples:
+//   # Run all benchmarks
+//   go test -bench=. -run='^$' -benchtime=1s pool_pubsub_bench_test.go
+//
+//   # Run only pool benchmarks
+//   go test -bench=BenchmarkPool -run='^$' pool_pubsub_bench_test.go
+//
+//   # Run only PubSub benchmarks
+//   go test -bench=BenchmarkPubSub -run='^$' pool_pubsub_bench_test.go
+//
+//   # Compare between branches
+//   git checkout branch1 && go test -bench=. -run='^$' pool_pubsub_bench_test.go > branch1.txt
+//   git checkout branch2 && go test -bench=. -run='^$' pool_pubsub_bench_test.go > branch2.txt
+//   benchcmp branch1.txt branch2.txt
+//
+//   # Run with memory profiling
+//   go test -bench=BenchmarkPoolGetPut -run='^$' -memprofile=mem.prof pool_pubsub_bench_test.go
+//
+//   # Run with CPU profiling
+//   go test -bench=BenchmarkPoolGetPut -run='^$' -cpuprofile=cpu.prof pool_pubsub_bench_test.go
+
+package redis_test
+
+import (
+	"context"
+	"fmt"
+	"net"
+	"sync"
+	"testing"
+	"time"
+
+	"github.com/redis/go-redis/v9"
+	"github.com/redis/go-redis/v9/internal/pool"
+)
+
+// dummyDialer creates a mock connection for benchmarking
+func dummyDialer(ctx context.Context) (net.Conn, error) {
+	return &dummyConn{}, nil
+}
+
+// dummyConn implements net.Conn for benchmarking
+type dummyConn struct{}
+
+func (c *dummyConn) Read(b []byte) (n int, err error)  { return len(b), nil }
+func (c *dummyConn) Write(b []byte) (n int, err error) { return len(b), nil }
+func (c *dummyConn) Close() error                      { return nil }
+func (c *dummyConn) LocalAddr() net.Addr               { return &net.TCPAddr{IP: net.IPv4(127, 0, 0, 1), Port: 6379} }
+func (c *dummyConn) RemoteAddr() net.Addr {
+	return &net.TCPAddr{IP: net.IPv4(127, 0, 0, 1), Port: 6379}
+}
+func (c *dummyConn) SetDeadline(t time.Time) error      { return nil }
+func (c *dummyConn) SetReadDeadline(t time.Time) error  { return nil }
+func (c *dummyConn) SetWriteDeadline(t time.Time) error { return nil }
+
+// =============================================================================
+// POOL BENCHMARKS
+// =============================================================================
+
+// BenchmarkPoolGetPut benchmarks the core pool Get/Put operations
+func BenchmarkPoolGetPut(b *testing.B) {
+	ctx := context.Background()
+
+	poolSizes := []int{1, 2, 4, 8, 16, 32, 64, 128}
+
+	for _, poolSize := range poolSizes {
+		b.Run(fmt.Sprintf("PoolSize_%d", poolSize), func(b *testing.B) {
+			connPool := pool.NewConnPool(&pool.Options{
+				Dialer:          dummyDialer,
+				PoolSize:        int32(poolSize),
+				PoolTimeout:     time.Second,
+				DialTimeout:     time.Second,
+				ConnMaxIdleTime: time.Hour,
+				MinIdleConns:    int32(0), // Start with no idle connections
+			})
+			defer connPool.Close()
+
+			b.ResetTimer()
+			b.ReportAllocs()
+
+			b.RunParallel(func(pb *testing.PB) {
+				for pb.Next() {
+					cn, err := connPool.Get(ctx)
+					if err != nil {
+						b.Fatal(err)
+					}
+					connPool.Put(ctx, cn)
+				}
+			})
+		})
+	}
+}
+
+// BenchmarkPoolGetPutWithMinIdle benchmarks pool operations with MinIdleConns
+func BenchmarkPoolGetPutWithMinIdle(b *testing.B) {
+	ctx := context.Background()
+
+	configs := []struct {
+		poolSize     int
+		minIdleConns int
+	}{
+		{8, 2},
+		{16, 4},
+		{32, 8},
+		{64, 16},
+	}
+
+	for _, config := range configs {
+		b.Run(fmt.Sprintf("Pool_%d_MinIdle_%d", config.poolSize, config.minIdleConns), func(b *testing.B) {
+			connPool := pool.NewConnPool(&pool.Options{
+				Dialer:          dummyDialer,
+				PoolSize:        int32(config.poolSize),
+				MinIdleConns:    int32(config.minIdleConns),
+				PoolTimeout:     time.Second,
+				DialTimeout:     time.Second,
+				ConnMaxIdleTime: time.Hour,
+			})
+			defer connPool.Close()
+
+			b.ResetTimer()
+			b.ReportAllocs()
+
+			b.RunParallel(func(pb *testing.PB) {
+				for pb.Next() {
+					cn, err := connPool.Get(ctx)
+					if err != nil {
+						b.Fatal(err)
+					}
+					connPool.Put(ctx, cn)
+				}
+			})
+		})
+	}
+}
+
+// BenchmarkPoolConcurrentGetPut benchmarks pool under high concurrency
+func BenchmarkPoolConcurrentGetPut(b *testing.B) {
+	ctx := context.Background()
+
+	connPool := pool.NewConnPool(&pool.Options{
+		Dialer:          dummyDialer,
+		PoolSize:        int32(32),
+		PoolTimeout:     time.Second,
+		DialTimeout:     time.Second,
+		ConnMaxIdleTime: time.Hour,
+		MinIdleConns:    int32(0),
+	})
+	defer connPool.Close()
+
+	b.ResetTimer()
+	b.ReportAllocs()
+
+	// Test with different levels of concurrency
+	concurrencyLevels := []int{1, 2, 4, 8, 16, 32, 64}
+
+	for _, concurrency := range concurrencyLevels {
+		b.Run(fmt.Sprintf("Concurrency_%d", concurrency), func(b *testing.B) {
+			b.SetParallelism(concurrency)
+			b.RunParallel(func(pb *testing.PB) {
+				for pb.Next() {
+					cn, err := connPool.Get(ctx)
+					if err != nil {
+						b.Fatal(err)
+					}
+					connPool.Put(ctx, cn)
+				}
+			})
+		})
+	}
+}
+
+// =============================================================================
+// PUBSUB BENCHMARKS
+// =============================================================================
+
+// benchmarkClient creates a Redis client for benchmarking with mock dialer
+func benchmarkClient(poolSize int) *redis.Client {
+	return redis.NewClient(&redis.Options{
+		Addr:         "localhost:6379", // Mock address
+		DialTimeout:  time.Second,
+		ReadTimeout:  time.Second,
+		WriteTimeout: time.Second,
+		PoolSize:     poolSize,
+		MinIdleConns: 0, // Start with no idle connections for consistent benchmarks
+	})
+}
+
+// BenchmarkPubSubCreation benchmarks PubSub creation and subscription
+func BenchmarkPubSubCreation(b *testing.B) {
+	ctx := context.Background()
+
+	poolSizes := []int{1, 4, 8, 16, 32}
+
+	for _, poolSize := range poolSizes {
+		b.Run(fmt.Sprintf("PoolSize_%d", poolSize), func(b *testing.B) {
+			client := benchmarkClient(poolSize)
+			defer client.Close()
+
+			b.ResetTimer()
+			b.ReportAllocs()
+
+			for i := 0; i < b.N; i++ {
+				pubsub := client.Subscribe(ctx, "test-channel")
+				pubsub.Close()
+			}
+		})
+	}
+}
+
+// BenchmarkPubSubPatternCreation benchmarks PubSub pattern subscription
+func BenchmarkPubSubPatternCreation(b *testing.B) {
+	ctx := context.Background()
+
+	poolSizes := []int{1, 4, 8, 16, 32}
+
+	for _, poolSize := range poolSizes {
+		b.Run(fmt.Sprintf("PoolSize_%d", poolSize), func(b *testing.B) {
+			client := benchmarkClient(poolSize)
+			defer client.Close()
+
+			b.ResetTimer()
+			b.ReportAllocs()
+
+			for i := 0; i < b.N; i++ {
+				pubsub := client.PSubscribe(ctx, "test-*")
+				pubsub.Close()
+			}
+		})
+	}
+}
+
+// BenchmarkPubSubConcurrentCreation benchmarks concurrent PubSub creation
+func BenchmarkPubSubConcurrentCreation(b *testing.B) {
+	ctx := context.Background()
+	client := benchmarkClient(32)
+	defer client.Close()
+
+	concurrencyLevels := []int{1, 2, 4, 8, 16}
+
+	for _, concurrency := range concurrencyLevels {
+		b.Run(fmt.Sprintf("Concurrency_%d", concurrency), func(b *testing.B) {
+			b.ResetTimer()
+			b.ReportAllocs()
+
+			var wg sync.WaitGroup
+			semaphore := make(chan struct{}, concurrency)
+
+			for i := 0; i < b.N; i++ {
+				wg.Add(1)
+				semaphore <- struct{}{}
+
+				go func() {
+					defer wg.Done()
+					defer func() { <-semaphore }()
+
+					pubsub := client.Subscribe(ctx, "test-channel")
+					pubsub.Close()
+				}()
+			}
+
+			wg.Wait()
+		})
+	}
+}
+
+// BenchmarkPubSubMultipleChannels benchmarks subscribing to multiple channels
+func BenchmarkPubSubMultipleChannels(b *testing.B) {
+	ctx := context.Background()
+	client := benchmarkClient(16)
+	defer client.Close()
+
+	channelCounts := []int{1, 5, 10, 25, 50, 100}
+
+	for _, channelCount := range channelCounts {
+		b.Run(fmt.Sprintf("Channels_%d", channelCount), func(b *testing.B) {
+			// Prepare channel names
+			channels := make([]string, channelCount)
+			for i := 0; i < channelCount; i++ {
+				channels[i] = fmt.Sprintf("channel-%d", i)
+			}
+
+			b.ResetTimer()
+			b.ReportAllocs()
+
+			for i := 0; i < b.N; i++ {
+				pubsub := client.Subscribe(ctx, channels...)
+				pubsub.Close()
+			}
+		})
+	}
+}
+
+// BenchmarkPubSubReuse benchmarks reusing PubSub connections
+func BenchmarkPubSubReuse(b *testing.B) {
+	ctx := context.Background()
+	client := benchmarkClient(16)
+	defer client.Close()
+
+	b.ResetTimer()
+	b.ReportAllocs()
+
+	for i := 0; i < b.N; i++ {
+		// Benchmark just the creation and closing of PubSub connections
+		// This simulates reuse patterns without requiring actual Redis operations
+		pubsub := client.Subscribe(ctx, fmt.Sprintf("test-channel-%d", i))
+		pubsub.Close()
+	}
+}
+
+// =============================================================================
+// COMBINED BENCHMARKS
+// =============================================================================
+
+// BenchmarkPoolAndPubSubMixed benchmarks mixed pool stats and PubSub operations
+func BenchmarkPoolAndPubSubMixed(b *testing.B) {
+	ctx := context.Background()
+	client := benchmarkClient(32)
+	defer client.Close()
+
+	b.ResetTimer()
+	b.ReportAllocs()
+
+	b.RunParallel(func(pb *testing.PB) {
+		for pb.Next() {
+			// Mix of pool stats collection and PubSub creation
+			if pb.Next() {
+				// Pool stats operation
+				stats := client.PoolStats()
+				_ = stats.Hits + stats.Misses // Use the stats to prevent optimization
+			}
+
+			if pb.Next() {
+				// PubSub operation
+				pubsub := client.Subscribe(ctx, "test-channel")
+				pubsub.Close()
+			}
+		}
+	})
+}
+
+// BenchmarkPoolStatsCollection benchmarks pool statistics collection
+func BenchmarkPoolStatsCollection(b *testing.B) {
+	client := benchmarkClient(16)
+	defer client.Close()
+
+	b.ResetTimer()
+	b.ReportAllocs()
+
+	for i := 0; i < b.N; i++ {
+		stats := client.PoolStats()
+		_ = stats.Hits + stats.Misses + stats.Timeouts // Use the stats to prevent optimization
+	}
+}
+
+// BenchmarkPoolHighContention tests pool performance under high contention
+func BenchmarkPoolHighContention(b *testing.B) {
+	ctx := context.Background()
+	client := benchmarkClient(32)
+	defer client.Close()
+
+	b.ResetTimer()
+	b.ReportAllocs()
+
+	b.RunParallel(func(pb *testing.PB) {
+		for pb.Next() {
+			// High contention Get/Put operations
+			pubsub := client.Subscribe(ctx, "test-channel")
+			pubsub.Close()
+		}
+	})
+}

pubsub.go

@@ -10,6 +10,7 @@ import (
 	"github.com/redis/go-redis/v9/internal"
 	"github.com/redis/go-redis/v9/internal/pool"
 	"github.com/redis/go-redis/v9/internal/proto"
+	"github.com/redis/go-redis/v9/push"
 )
 
 // PubSub implements Pub/Sub commands as described in
@@ -21,7 +22,7 @@ import (
 type PubSub struct {
 	opt *Options
 
-	newConn   func(ctx context.Context, channels []string) (*pool.Conn, error)
+	newConn   func(ctx context.Context, addr string, channels []string) (*pool.Conn, error)
 	closeConn func(*pool.Conn) error
 
 	mu        sync.Mutex
@@ -38,6 +39,12 @@ type PubSub struct {
 	chOnce sync.Once
 	msgCh  *channel
 	allCh  *channel
+
+	// Push notification processor for handling generic push notifications
+	pushProcessor push.NotificationProcessor
+
+	// Cleanup callback for hitless upgrade tracking
+	onClose func()
 }
 
 func (c *PubSub) init() {
@@ -69,10 +76,18 @@ func (c *PubSub) conn(ctx context.Context, newChannels []string) (*pool.Conn, er
 		return c.cn, nil
 	}
 
+	if c.opt.Addr == "" {
+		// TODO(hitless):
+		// this is probably cluster client
+		// c.newConn will ignore the addr argument
+		// will be changed when we have hitless upgrades for cluster clients
+		c.opt.Addr = internal.RedisNull
+	}
+
 	channels := mapKeys(c.channels)
 	channels = append(channels, newChannels...)
 
-	cn, err := c.newConn(ctx, channels)
+	cn, err := c.newConn(ctx, c.opt.Addr, channels)
 	if err != nil {
 		return nil, err
 	}
@@ -153,12 +168,31 @@ func (c *PubSub) releaseConn(ctx context.Context, cn *pool.Conn, err error, allo
 	if c.cn != cn {
 		return
 	}
+
+	if !cn.IsUsable() || cn.ShouldHandoff() {
+		c.reconnect(ctx, fmt.Errorf("pubsub: connection is not usable"))
+	}
+
 	if isBadConn(err, allowTimeout, c.opt.Addr) {
 		c.reconnect(ctx, err)
 	}
 }
 
 func (c *PubSub) reconnect(ctx context.Context, reason error) {
+	if c.cn != nil && c.cn.ShouldHandoff() {
+		newEndpoint := c.cn.GetHandoffEndpoint()
+		// If new endpoint is NULL, use the original address
+		if newEndpoint == internal.RedisNull {
+			newEndpoint = c.opt.Addr
+		}
+
+		if newEndpoint != "" {
+			// Update the address in the options
+			oldAddr := c.cn.RemoteAddr().String()
+			c.opt.Addr = newEndpoint
+			internal.Logger.Printf(ctx, "pubsub: reconnecting to new endpoint %s (was %s)", newEndpoint, oldAddr)
+		}
+	}
 	_ = c.closeTheCn(reason)
 	_, _ = c.conn(ctx, nil)
 }
@@ -167,9 +201,6 @@ func (c *PubSub) closeTheCn(reason error) error {
 	if c.cn == nil {
 		return nil
 	}
-	if !c.closed {
-		internal.Logger.Printf(c.getContext(), "redis: discarding bad PubSub connection: %s", reason)
-	}
 	err := c.closeConn(c.cn)
 	c.cn = nil
 	return err
@@ -185,6 +216,11 @@ func (c *PubSub) Close() error {
 	c.closed = true
 	close(c.exit)
 
+	// Call cleanup callback if set
+	if c.onClose != nil {
+		c.onClose()
+	}
+
 	return c.closeTheCn(pool.ErrClosed)
 }
 
@@ -436,9 +472,14 @@ func (c *PubSub) ReceiveTimeout(ctx context.Context, timeout time.Duration) (int
 	}
 
 	err = cn.WithReader(ctx, timeout, func(rd *proto.Reader) error {
+		// To be sure there are no buffered push notifications, we process them before reading the reply
+		if err := c.processPendingPushNotificationWithReader(ctx, cn, rd); err != nil {
+			// Log the error but don't fail the command execution
+			// Push notification processing errors shouldn't break normal Redis operations
+			internal.Logger.Printf(ctx, "push: conn[%d] error processing pending notifications before reading reply: %v", cn.GetID(), err)
+		}
 		return c.cmd.readReply(rd)
 	})
-
 	c.releaseConnWithLock(ctx, cn, err, timeout > 0)
 
 	if err != nil {
@@ -451,6 +492,12 @@ func (c *PubSub) ReceiveTimeout(ctx context.Context, timeout time.Duration) (int
 // Receive returns a message as a Subscription, Message, Pong or error.
 // See PubSub example for details. This is low-level API and in most cases
 // Channel should be used instead.
+// Receive returns a message as a Subscription, Message, Pong, or an error.
+// See PubSub example for details. This is a low-level API and in most cases
+// Channel should be used instead.
+// This method blocks until a message is received or an error occurs.
+// It may return early with an error if the context is canceled, the connection fails,
+// or other internal errors occur.
 func (c *PubSub) Receive(ctx context.Context) (interface{}, error) {
 	return c.ReceiveTimeout(ctx, 0)
 }
@@ -532,6 +579,27 @@ func (c *PubSub) ChannelWithSubscriptions(opts ...ChannelOption) <-chan interfac
 	return c.allCh.allCh
 }
 
+func (c *PubSub) processPendingPushNotificationWithReader(ctx context.Context, cn *pool.Conn, rd *proto.Reader) error {
+	// Only process push notifications for RESP3 connections with a processor
+	if c.opt.Protocol != 3 || c.pushProcessor == nil {
+		return nil
+	}
+
+	// Create handler context with client, connection pool, and connection information
+	handlerCtx := c.pushNotificationHandlerContext(cn)
+	return c.pushProcessor.ProcessPendingNotifications(ctx, handlerCtx, rd)
+}
+
+func (c *PubSub) pushNotificationHandlerContext(cn *pool.Conn) push.NotificationHandlerContext {
+	// PubSub doesn't have a client or connection pool, so we pass nil for those
+	// PubSub connections are blocking
+	return push.NotificationHandlerContext{
+		PubSub:     c,
+		Conn:       cn,
+		IsBlocking: true,
+	}
+}
+
 type ChannelOption func(c *channel)
 
 // WithChannelSize specifies the Go chan size that is used to buffer incoming messages.

push/errors.go

@@ -0,0 +1,170 @@
+package push
+
+import (
+	"errors"
+	"fmt"
+)
+
+// Push notification error definitions
+// This file contains all error types and messages used by the push notification system
+
+// Error reason constants
+const (
+	// HandlerReasons
+	ReasonHandlerNil       = "handler cannot be nil"
+	ReasonHandlerExists    = "cannot overwrite existing handler"
+	ReasonHandlerProtected = "handler is protected"
+
+	// ProcessorReasons
+	ReasonPushNotificationsDisabled = "push notifications are disabled"
+)
+
+// ProcessorType represents the type of processor involved in the error
+// defined as a custom type for better readability and easier maintenance
+type ProcessorType string
+
+const (
+	// ProcessorTypes
+	ProcessorTypeProcessor     = ProcessorType("processor")
+	ProcessorTypeVoidProcessor = ProcessorType("void_processor")
+	ProcessorTypeCustom        = ProcessorType("custom")
+)
+
+// ProcessorOperation represents the operation being performed by the processor
+// defined as a custom type for better readability and easier maintenance
+type ProcessorOperation string
+
+const (
+	// ProcessorOperations
+	ProcessorOperationProcess    = ProcessorOperation("process")
+	ProcessorOperationRegister   = ProcessorOperation("register")
+	ProcessorOperationUnregister = ProcessorOperation("unregister")
+	ProcessorOperationUnknown    = ProcessorOperation("unknown")
+)
+
+// Common error variables for reuse
+var (
+	// ErrHandlerNil is returned when attempting to register a nil handler
+	ErrHandlerNil = errors.New(ReasonHandlerNil)
+)
+
+// Registry errors
+
+// ErrHandlerExists creates an error for when attempting to overwrite an existing handler
+func ErrHandlerExists(pushNotificationName string) error {
+	return NewHandlerError(ProcessorOperationRegister, pushNotificationName, ReasonHandlerExists, nil)
+}
+
+// ErrProtectedHandler creates an error for when attempting to unregister a protected handler
+func ErrProtectedHandler(pushNotificationName string) error {
+	return NewHandlerError(ProcessorOperationUnregister, pushNotificationName, ReasonHandlerProtected, nil)
+}
+
+// VoidProcessor errors
+
+// ErrVoidProcessorRegister creates an error for when attempting to register a handler on void processor
+func ErrVoidProcessorRegister(pushNotificationName string) error {
+	return NewProcessorError(ProcessorTypeVoidProcessor, ProcessorOperationRegister, pushNotificationName, ReasonPushNotificationsDisabled, nil)
+}
+
+// ErrVoidProcessorUnregister creates an error for when attempting to unregister a handler on void processor
+func ErrVoidProcessorUnregister(pushNotificationName string) error {
+	return NewProcessorError(ProcessorTypeVoidProcessor, ProcessorOperationUnregister, pushNotificationName, ReasonPushNotificationsDisabled, nil)
+}
+
+// Error type definitions for advanced error handling
+
+// HandlerError represents errors related to handler operations
+type HandlerError struct {
+	Operation            ProcessorOperation
+	PushNotificationName string
+	Reason               string
+	Err                  error
+}
+
+func (e *HandlerError) Error() string {
+	if e.Err != nil {
+		return fmt.Sprintf("handler %s failed for '%s': %s (%v)", e.Operation, e.PushNotificationName, e.Reason, e.Err)
+	}
+	return fmt.Sprintf("handler %s failed for '%s': %s", e.Operation, e.PushNotificationName, e.Reason)
+}
+
+func (e *HandlerError) Unwrap() error {
+	return e.Err
+}
+
+// NewHandlerError creates a new HandlerError
+func NewHandlerError(operation ProcessorOperation, pushNotificationName, reason string, err error) *HandlerError {
+	return &HandlerError{
+		Operation:            operation,
+		PushNotificationName: pushNotificationName,
+		Reason:               reason,
+		Err:                  err,
+	}
+}
+
+// ProcessorError represents errors related to processor operations
+type ProcessorError struct {
+	ProcessorType        ProcessorType      // "processor", "void_processor"
+	Operation            ProcessorOperation // "process", "register", "unregister"
+	PushNotificationName string             // Name of the push notification involved
+	Reason               string
+	Err                  error
+}
+
+func (e *ProcessorError) Error() string {
+	notifInfo := ""
+	if e.PushNotificationName != "" {
+		notifInfo = fmt.Sprintf(" for '%s'", e.PushNotificationName)
+	}
+	if e.Err != nil {
+		return fmt.Sprintf("%s %s failed%s: %s (%v)", e.ProcessorType, e.Operation, notifInfo, e.Reason, e.Err)
+	}
+	return fmt.Sprintf("%s %s failed%s: %s", e.ProcessorType, e.Operation, notifInfo, e.Reason)
+}
+
+func (e *ProcessorError) Unwrap() error {
+	return e.Err
+}
+
+// NewProcessorError creates a new ProcessorError
+func NewProcessorError(processorType ProcessorType, operation ProcessorOperation, pushNotificationName, reason string, err error) *ProcessorError {
+	return &ProcessorError{
+		ProcessorType:        processorType,
+		Operation:            operation,
+		PushNotificationName: pushNotificationName,
+		Reason:               reason,
+		Err:                  err,
+	}
+}
+
+// Helper functions for common error scenarios
+
+// IsHandlerNilError checks if an error is due to a nil handler
+func IsHandlerNilError(err error) bool {
+	return errors.Is(err, ErrHandlerNil)
+}
+
+// IsHandlerExistsError checks if an error is due to attempting to overwrite an existing handler
+func IsHandlerExistsError(err error) bool {
+	if handlerErr, ok := err.(*HandlerError); ok {
+		return handlerErr.Operation == ProcessorOperationRegister && handlerErr.Reason == ReasonHandlerExists
+	}
+	return false
+}
+
+// IsProtectedHandlerError checks if an error is due to attempting to unregister a protected handler
+func IsProtectedHandlerError(err error) bool {
+	if handlerErr, ok := err.(*HandlerError); ok {
+		return handlerErr.Operation == ProcessorOperationUnregister && handlerErr.Reason == ReasonHandlerProtected
+	}
+	return false
+}
+
+// IsVoidProcessorError checks if an error is due to void processor operations
+func IsVoidProcessorError(err error) bool {
+	if procErr, ok := err.(*ProcessorError); ok {
+		return procErr.ProcessorType == ProcessorTypeVoidProcessor && procErr.Reason == ReasonPushNotificationsDisabled
+	}
+	return false
+}

push/handler.go

@@ -0,0 +1,14 @@
+package push
+
+import (
+	"context"
+)
+
+// NotificationHandler defines the interface for push notification handlers.
+type NotificationHandler interface {
+	// HandlePushNotification processes a push notification with context information.
+	// The handlerCtx provides information about the client, connection pool, and connection
+	// on which the notification was received, allowing handlers to make informed decisions.
+	// Returns an error if the notification could not be handled.
+	HandlePushNotification(ctx context.Context, handlerCtx NotificationHandlerContext, notification []interface{}) error
+}

push/handler_context.go

@@ -0,0 +1,44 @@
+package push
+
+// No imports needed for this file
+
+// NotificationHandlerContext provides context information about where a push notification was received.
+// This struct allows handlers to make informed decisions based on the source of the notification
+// with strongly typed access to different client types using concrete types.
+type NotificationHandlerContext struct {
+	// Client is the Redis client instance that received the notification.
+	// It is interface to both allow for future expansion and to avoid
+	// circular dependencies. The developer is responsible for type assertion.
+	// It can be one of the following types:
+	// - *redis.baseClient
+	// - *redis.Client
+	// - *redis.ClusterClient
+	// - *redis.Conn
+	Client interface{}
+
+	// ConnPool is the connection pool from which the connection was obtained.
+	// It is interface to both allow for future expansion and to avoid
+	// circular dependencies. The developer is responsible for type assertion.
+	// It can be one of the following types:
+	// - *pool.ConnPool
+	// - *pool.SingleConnPool
+	// - *pool.StickyConnPool
+	ConnPool interface{}
+
+	// PubSub is the PubSub instance that received the notification.
+	// It is interface to both allow for future expansion and to avoid
+	// circular dependencies. The developer is responsible for type assertion.
+	// It can be one of the following types:
+	// - *redis.PubSub
+	PubSub interface{}
+
+	// Conn is the specific connection on which the notification was received.
+	// It is interface to both allow for future expansion and to avoid
+	// circular dependencies. The developer is responsible for type assertion.
+	// It can be one of the following types:
+	// - *pool.Conn
+	Conn interface{}
+
+	// IsBlocking indicates if the notification was received on a blocking connection.
+	IsBlocking bool
+}

push/processor.go

@@ -0,0 +1,203 @@
+package push
+
+import (
+	"context"
+
+	"github.com/redis/go-redis/v9/internal"
+	"github.com/redis/go-redis/v9/internal/proto"
+)
+
+// NotificationProcessor defines the interface for push notification processors.
+type NotificationProcessor interface {
+	// GetHandler returns the handler for a specific push notification name.
+	GetHandler(pushNotificationName string) NotificationHandler
+	// ProcessPendingNotifications checks for and processes any pending push notifications.
+	// To be used when it is known that there are notifications on the socket.
+	// It will try to read from the socket and if it is empty - it may block.
+	ProcessPendingNotifications(ctx context.Context, handlerCtx NotificationHandlerContext, rd *proto.Reader) error
+	// RegisterHandler registers a handler for a specific push notification name.
+	RegisterHandler(pushNotificationName string, handler NotificationHandler, protected bool) error
+	// UnregisterHandler removes a handler for a specific push notification name.
+	UnregisterHandler(pushNotificationName string) error
+}
+
+// Processor handles push notifications with a registry of handlers
+type Processor struct {
+	registry *Registry
+}
+
+// NewProcessor creates a new push notification processor
+func NewProcessor() *Processor {
+	return &Processor{
+		registry: NewRegistry(),
+	}
+}
+
+// GetHandler returns the handler for a specific push notification name
+func (p *Processor) GetHandler(pushNotificationName string) NotificationHandler {
+	return p.registry.GetHandler(pushNotificationName)
+}
+
+// RegisterHandler registers a handler for a specific push notification name
+func (p *Processor) RegisterHandler(pushNotificationName string, handler NotificationHandler, protected bool) error {
+	return p.registry.RegisterHandler(pushNotificationName, handler, protected)
+}
+
+// UnregisterHandler removes a handler for a specific push notification name
+func (p *Processor) UnregisterHandler(pushNotificationName string) error {
+	return p.registry.UnregisterHandler(pushNotificationName)
+}
+
+// ProcessPendingNotifications checks for and processes any pending push notifications
+// This method should be called by the client in WithReader before reading the reply
+// It will try to read from the socket and if it is empty - it may block.
+func (p *Processor) ProcessPendingNotifications(ctx context.Context, handlerCtx NotificationHandlerContext, rd *proto.Reader) error {
+	if rd == nil {
+		return nil
+	}
+
+	for {
+		// Check if there's data available to read
+		replyType, err := rd.PeekReplyType()
+		if err != nil {
+			// No more data available or error reading
+			// if timeout, it will be handled by the caller
+			break
+		}
+
+		// Only process push notifications (arrays starting with >)
+		if replyType != proto.RespPush {
+			break
+		}
+
+		// see if we should skip this notification
+		notificationName, err := rd.PeekPushNotificationName()
+		if err != nil {
+			break
+		}
+
+		if willHandleNotificationInClient(notificationName) {
+			break
+		}
+
+		// Read the push notification
+		reply, err := rd.ReadReply()
+		if err != nil {
+			internal.Logger.Printf(ctx, "push: error reading push notification: %v", err)
+			break
+		}
+
+		// Convert to slice of interfaces
+		notification, ok := reply.([]interface{})
+		if !ok {
+			break
+		}
+
+		// Handle the notification directly
+		if len(notification) > 0 {
+			// Extract the notification type (first element)
+			if notificationType, ok := notification[0].(string); ok {
+				// Get the handler for this notification type
+				if handler := p.registry.GetHandler(notificationType); handler != nil {
+					// Handle the notification
+					err := handler.HandlePushNotification(ctx, handlerCtx, notification)
+					if err != nil {
+						internal.Logger.Printf(ctx, "push: error handling push notification: %v", err)
+					}
+				}
+			}
+		}
+	}
+
+	return nil
+}
+
+// VoidProcessor discards all push notifications without processing them
+type VoidProcessor struct{}
+
+// NewVoidProcessor creates a new void push notification processor
+func NewVoidProcessor() *VoidProcessor {
+	return &VoidProcessor{}
+}
+
+// GetHandler returns nil for void processor since it doesn't maintain handlers
+func (v *VoidProcessor) GetHandler(_ string) NotificationHandler {
+	return nil
+}
+
+// RegisterHandler returns an error for void processor since it doesn't maintain handlers
+func (v *VoidProcessor) RegisterHandler(pushNotificationName string, _ NotificationHandler, _ bool) error {
+	return ErrVoidProcessorRegister(pushNotificationName)
+}
+
+// UnregisterHandler returns an error for void processor since it doesn't maintain handlers
+func (v *VoidProcessor) UnregisterHandler(pushNotificationName string) error {
+	return ErrVoidProcessorUnregister(pushNotificationName)
+}
+
+// ProcessPendingNotifications for VoidProcessor does nothing since push notifications
+// are only available in RESP3 and this processor is used for RESP2 connections.
+// This avoids unnecessary buffer scanning overhead.
+// It does however read and discard all push notifications from the buffer to avoid
+// them being interpreted as a reply.
+// This method should be called by the client in WithReader before reading the reply
+// to be sure there are no buffered push notifications.
+// It will try to read from the socket and if it is empty - it may block.
+func (v *VoidProcessor) ProcessPendingNotifications(_ context.Context, handlerCtx NotificationHandlerContext, rd *proto.Reader) error {
+	// read and discard all push notifications
+	if rd == nil {
+		return nil
+	}
+
+	for {
+		// Check if there's data available to read
+		replyType, err := rd.PeekReplyType()
+		if err != nil {
+			// No more data available or error reading
+			// if timeout, it will be handled by the caller
+			break
+		}
+
+		// Only process push notifications (arrays starting with >)
+		if replyType != proto.RespPush {
+			break
+		}
+		// see if we should skip this notification
+		notificationName, err := rd.PeekPushNotificationName()
+		if err != nil {
+			break
+		}
+
+		if willHandleNotificationInClient(notificationName) {
+			break
+		}
+
+		// Read the push notification
+		_, err = rd.ReadReply()
+		if err != nil {
+			internal.Logger.Printf(context.Background(), "push: error reading push notification: %v", err)
+			return nil
+		}
+	}
+	return nil
+}
+
+// willHandleNotificationInClient checks if a notification type should be ignored by the push notification
+// processor and handled by other specialized systems instead (pub/sub, streams, keyspace, etc.).
+func willHandleNotificationInClient(notificationType string) bool {
+	switch notificationType {
+	// Pub/Sub notifications - handled by pub/sub system
+	case "message", // Regular pub/sub message
+		"pmessage",     // Pattern pub/sub message
+		"subscribe",    // Subscription confirmation
+		"unsubscribe",  // Unsubscription confirmation
+		"psubscribe",   // Pattern subscription confirmation
+		"punsubscribe", // Pattern unsubscription confirmation
+		"smessage",     // Sharded pub/sub message (Redis 7.0+)
+		"ssubscribe",   // Sharded subscription confirmation
+		"sunsubscribe": // Sharded unsubscription confirmation
+		return true
+	default:
+		return false
+	}
+}

push/processor_unit_test.go

@@ -0,0 +1,315 @@
+package push
+
+import (
+	"context"
+	"testing"
+)
+
+// TestProcessorCreation tests processor creation and initialization
+func TestProcessorCreation(t *testing.T) {
+	t.Run("NewProcessor", func(t *testing.T) {
+		processor := NewProcessor()
+		if processor == nil {
+			t.Fatal("NewProcessor should not return nil")
+		}
+		if processor.registry == nil {
+			t.Error("Processor should have a registry")
+		}
+	})
+
+	t.Run("NewVoidProcessor", func(t *testing.T) {
+		voidProcessor := NewVoidProcessor()
+		if voidProcessor == nil {
+			t.Fatal("NewVoidProcessor should not return nil")
+		}
+	})
+}
+
+// TestProcessorHandlerManagement tests handler registration and retrieval
+func TestProcessorHandlerManagement(t *testing.T) {
+	processor := NewProcessor()
+	handler := &UnitTestHandler{name: "test-handler"}
+
+	t.Run("RegisterHandler", func(t *testing.T) {
+		err := processor.RegisterHandler("TEST", handler, false)
+		if err != nil {
+			t.Errorf("RegisterHandler should not error: %v", err)
+		}
+
+		// Verify handler is registered
+		retrievedHandler := processor.GetHandler("TEST")
+		if retrievedHandler != handler {
+			t.Error("GetHandler should return the registered handler")
+		}
+	})
+
+	t.Run("RegisterProtectedHandler", func(t *testing.T) {
+		protectedHandler := &UnitTestHandler{name: "protected-handler"}
+		err := processor.RegisterHandler("PROTECTED", protectedHandler, true)
+		if err != nil {
+			t.Errorf("RegisterHandler should not error for protected handler: %v", err)
+		}
+
+		// Verify handler is registered
+		retrievedHandler := processor.GetHandler("PROTECTED")
+		if retrievedHandler != protectedHandler {
+			t.Error("GetHandler should return the protected handler")
+		}
+	})
+
+	t.Run("GetNonExistentHandler", func(t *testing.T) {
+		handler := processor.GetHandler("NONEXISTENT")
+		if handler != nil {
+			t.Error("GetHandler should return nil for non-existent handler")
+		}
+	})
+
+	t.Run("UnregisterHandler", func(t *testing.T) {
+		err := processor.UnregisterHandler("TEST")
+		if err != nil {
+			t.Errorf("UnregisterHandler should not error: %v", err)
+		}
+
+		// Verify handler is removed
+		retrievedHandler := processor.GetHandler("TEST")
+		if retrievedHandler != nil {
+			t.Error("GetHandler should return nil after unregistering")
+		}
+	})
+
+	t.Run("UnregisterProtectedHandler", func(t *testing.T) {
+		err := processor.UnregisterHandler("PROTECTED")
+		if err == nil {
+			t.Error("UnregisterHandler should error for protected handler")
+		}
+
+		// Verify handler is still there
+		retrievedHandler := processor.GetHandler("PROTECTED")
+		if retrievedHandler == nil {
+			t.Error("Protected handler should not be removed")
+		}
+	})
+}
+
+// TestVoidProcessorBehavior tests void processor behavior
+func TestVoidProcessorBehavior(t *testing.T) {
+	voidProcessor := NewVoidProcessor()
+	handler := &UnitTestHandler{name: "test-handler"}
+
+	t.Run("GetHandler", func(t *testing.T) {
+		retrievedHandler := voidProcessor.GetHandler("ANY")
+		if retrievedHandler != nil {
+			t.Error("VoidProcessor GetHandler should always return nil")
+		}
+	})
+
+	t.Run("RegisterHandler", func(t *testing.T) {
+		err := voidProcessor.RegisterHandler("TEST", handler, false)
+		if err == nil {
+			t.Error("VoidProcessor RegisterHandler should return error")
+		}
+
+		// Check error type
+		if !IsVoidProcessorError(err) {
+			t.Error("Error should be a VoidProcessorError")
+		}
+	})
+
+	t.Run("UnregisterHandler", func(t *testing.T) {
+		err := voidProcessor.UnregisterHandler("TEST")
+		if err == nil {
+			t.Error("VoidProcessor UnregisterHandler should return error")
+		}
+
+		// Check error type
+		if !IsVoidProcessorError(err) {
+			t.Error("Error should be a VoidProcessorError")
+		}
+	})
+}
+
+// TestProcessPendingNotificationsNilReader tests handling of nil reader
+func TestProcessPendingNotificationsNilReader(t *testing.T) {
+	t.Run("ProcessorWithNilReader", func(t *testing.T) {
+		processor := NewProcessor()
+		ctx := context.Background()
+		handlerCtx := NotificationHandlerContext{}
+
+		err := processor.ProcessPendingNotifications(ctx, handlerCtx, nil)
+		if err != nil {
+			t.Errorf("ProcessPendingNotifications should not error with nil reader: %v", err)
+		}
+	})
+
+	t.Run("VoidProcessorWithNilReader", func(t *testing.T) {
+		voidProcessor := NewVoidProcessor()
+		ctx := context.Background()
+		handlerCtx := NotificationHandlerContext{}
+
+		err := voidProcessor.ProcessPendingNotifications(ctx, handlerCtx, nil)
+		if err != nil {
+			t.Errorf("VoidProcessor ProcessPendingNotifications should not error with nil reader: %v", err)
+		}
+	})
+}
+
+// TestWillHandleNotificationInClient tests the notification filtering logic
+func TestWillHandleNotificationInClient(t *testing.T) {
+	testCases := []struct {
+		name           string
+		notificationType string
+		shouldHandle   bool
+	}{
+		// Pub/Sub notifications (should be handled in client)
+		{"message", "message", true},
+		{"pmessage", "pmessage", true},
+		{"subscribe", "subscribe", true},
+		{"unsubscribe", "unsubscribe", true},
+		{"psubscribe", "psubscribe", true},
+		{"punsubscribe", "punsubscribe", true},
+		{"smessage", "smessage", true},
+		{"ssubscribe", "ssubscribe", true},
+		{"sunsubscribe", "sunsubscribe", true},
+
+		// Push notifications (should be handled by processor)
+		{"MOVING", "MOVING", false},
+		{"MIGRATING", "MIGRATING", false},
+		{"MIGRATED", "MIGRATED", false},
+		{"FAILING_OVER", "FAILING_OVER", false},
+		{"FAILED_OVER", "FAILED_OVER", false},
+		{"custom", "custom", false},
+		{"unknown", "unknown", false},
+		{"empty", "", false},
+	}
+
+	for _, tc := range testCases {
+		t.Run(tc.name, func(t *testing.T) {
+			result := willHandleNotificationInClient(tc.notificationType)
+			if result != tc.shouldHandle {
+				t.Errorf("willHandleNotificationInClient(%q) = %v, want %v", tc.notificationType, result, tc.shouldHandle)
+			}
+		})
+	}
+}
+
+// TestProcessorErrorHandlingUnit tests error handling scenarios
+func TestProcessorErrorHandlingUnit(t *testing.T) {
+	processor := NewProcessor()
+
+	t.Run("RegisterNilHandler", func(t *testing.T) {
+		err := processor.RegisterHandler("TEST", nil, false)
+		if err == nil {
+			t.Error("RegisterHandler should error with nil handler")
+		}
+
+		// Check error type
+		if !IsHandlerNilError(err) {
+			t.Error("Error should be a HandlerNilError")
+		}
+	})
+
+	t.Run("RegisterDuplicateHandler", func(t *testing.T) {
+		handler1 := &UnitTestHandler{name: "handler1"}
+		handler2 := &UnitTestHandler{name: "handler2"}
+
+		// Register first handler
+		err := processor.RegisterHandler("DUPLICATE", handler1, false)
+		if err != nil {
+			t.Errorf("First RegisterHandler should not error: %v", err)
+		}
+
+		// Try to register second handler with same name
+		err = processor.RegisterHandler("DUPLICATE", handler2, false)
+		if err == nil {
+			t.Error("RegisterHandler should error when registering duplicate handler")
+		}
+
+		// Verify original handler is still there
+		retrievedHandler := processor.GetHandler("DUPLICATE")
+		if retrievedHandler != handler1 {
+			t.Error("Original handler should remain after failed duplicate registration")
+		}
+	})
+
+	t.Run("UnregisterNonExistentHandler", func(t *testing.T) {
+		err := processor.UnregisterHandler("NONEXISTENT")
+		if err != nil {
+			t.Errorf("UnregisterHandler should not error for non-existent handler: %v", err)
+		}
+	})
+}
+
+// TestProcessorConcurrentAccess tests concurrent access to processor
+func TestProcessorConcurrentAccess(t *testing.T) {
+	processor := NewProcessor()
+	
+	t.Run("ConcurrentRegisterAndGet", func(t *testing.T) {
+		done := make(chan bool, 2)
+
+		// Goroutine 1: Register handlers
+		go func() {
+			defer func() { done <- true }()
+			for i := 0; i < 100; i++ {
+				handler := &UnitTestHandler{name: "concurrent-handler"}
+				processor.RegisterHandler("CONCURRENT", handler, false)
+				processor.UnregisterHandler("CONCURRENT")
+			}
+		}()
+
+		// Goroutine 2: Get handlers
+		go func() {
+			defer func() { done <- true }()
+			for i := 0; i < 100; i++ {
+				processor.GetHandler("CONCURRENT")
+			}
+		}()
+
+		// Wait for both goroutines to complete
+		<-done
+		<-done
+	})
+}
+
+// TestProcessorInterfaceCompliance tests interface compliance
+func TestProcessorInterfaceCompliance(t *testing.T) {
+	t.Run("ProcessorImplementsInterface", func(t *testing.T) {
+		var _ NotificationProcessor = (*Processor)(nil)
+	})
+
+	t.Run("VoidProcessorImplementsInterface", func(t *testing.T) {
+		var _ NotificationProcessor = (*VoidProcessor)(nil)
+	})
+}
+
+// UnitTestHandler is a test implementation of NotificationHandler
+type UnitTestHandler struct {
+	name           string
+	lastNotification []interface{}
+	errorToReturn  error
+	callCount      int
+}
+
+func (h *UnitTestHandler) HandlePushNotification(ctx context.Context, handlerCtx NotificationHandlerContext, notification []interface{}) error {
+	h.callCount++
+	h.lastNotification = notification
+	return h.errorToReturn
+}
+
+// Helper methods for UnitTestHandler
+func (h *UnitTestHandler) GetCallCount() int {
+	return h.callCount
+}
+
+func (h *UnitTestHandler) GetLastNotification() []interface{} {
+	return h.lastNotification
+}
+
+func (h *UnitTestHandler) SetErrorToReturn(err error) {
+	h.errorToReturn = err
+}
+
+func (h *UnitTestHandler) Reset() {
+	h.callCount = 0
+	h.lastNotification = nil
+	h.errorToReturn = nil
+}

push/push.go

@@ -0,0 +1,7 @@
+// Package push provides push notifications for Redis.
+// This is an EXPERIMENTAL API for handling push notifications from Redis.
+// It is not yet stable and may change in the future.
+// Although this is in a public package, in its current form public use is not advised.
+// Pending push notifications should be processed before executing any readReply from the connection
+// as per RESP3 specification push notifications can be sent at any time.
+package push

push/registry.go

@@ -0,0 +1,61 @@
+package push
+
+import (
+	"sync"
+)
+
+// Registry manages push notification handlers
+type Registry struct {
+	mu        sync.RWMutex
+	handlers  map[string]NotificationHandler
+	protected map[string]bool
+}
+
+// NewRegistry creates a new push notification registry
+func NewRegistry() *Registry {
+	return &Registry{
+		handlers:  make(map[string]NotificationHandler),
+		protected: make(map[string]bool),
+	}
+}
+
+// RegisterHandler registers a handler for a specific push notification name
+func (r *Registry) RegisterHandler(pushNotificationName string, handler NotificationHandler, protected bool) error {
+	if handler == nil {
+		return ErrHandlerNil
+	}
+
+	r.mu.Lock()
+	defer r.mu.Unlock()
+
+	// Check if handler already exists
+	if _, exists := r.protected[pushNotificationName]; exists {
+		return ErrHandlerExists(pushNotificationName)
+	}
+
+	r.handlers[pushNotificationName] = handler
+	r.protected[pushNotificationName] = protected
+	return nil
+}
+
+// GetHandler returns the handler for a specific push notification name
+func (r *Registry) GetHandler(pushNotificationName string) NotificationHandler {
+	r.mu.RLock()
+	defer r.mu.RUnlock()
+	return r.handlers[pushNotificationName]
+}
+
+// UnregisterHandler removes a handler for a specific push notification name
+func (r *Registry) UnregisterHandler(pushNotificationName string) error {
+	r.mu.Lock()
+	defer r.mu.Unlock()
+
+	// Check if handler is protected
+	if protected, exists := r.protected[pushNotificationName]; exists && protected {
+		return ErrProtectedHandler(pushNotificationName)
+	}
+
+	delete(r.handlers, pushNotificationName)
+	delete(r.protected, pushNotificationName)
+	return nil
+}

push_notifications.go

@@ -0,0 +1,21 @@
+package redis
+
+import (
+	"github.com/redis/go-redis/v9/push"
+)
+
+// NewPushNotificationProcessor creates a new push notification processor
+// This processor maintains a registry of handlers and processes push notifications
+// It is used for RESP3 connections where push notifications are available
+func NewPushNotificationProcessor() push.NotificationProcessor {
+	return push.NewProcessor()
+}
+
+// NewVoidPushNotificationProcessor creates a new void push notification processor
+// This processor does not maintain any handlers and always returns nil for all operations
+// It is used for RESP2 connections where push notifications are not available
+// It can also be used to disable push notifications for RESP3 connections, where
+// it will discard all push notifications without processing them
+func NewVoidPushNotificationProcessor() push.NotificationProcessor {
+	return push.NewVoidProcessor()
+}

redis.go

@@ -10,10 +10,12 @@ import (
 	"time"
 
 	"github.com/redis/go-redis/v9/auth"
+	"github.com/redis/go-redis/v9/hitless"
 	"github.com/redis/go-redis/v9/internal"
 	"github.com/redis/go-redis/v9/internal/hscan"
 	"github.com/redis/go-redis/v9/internal/pool"
 	"github.com/redis/go-redis/v9/internal/proto"
+	"github.com/redis/go-redis/v9/push"
 )
 
 // Scanner internal/hscan.Scanner exposed interface.
@@ -23,6 +25,7 @@ type Scanner = hscan.Scanner
 const Nil = proto.Nil
 
 // SetLogger set custom log
+// Use with VoidLogger to disable logging.
 func SetLogger(logger internal.Logging) {
 	internal.Logger = logger
 }
@@ -203,15 +206,34 @@ func (hs *hooksMixin) processTxPipelineHook(ctx context.Context, cmds []Cmder) e
 
 type baseClient struct {
 	opt        *Options
+	optLock    sync.RWMutex
 	connPool   pool.Pooler
+	pubSubPool *pool.PubSubPool
 	hooksMixin
 
 	onClose func() error // hook called when client is closed
+
+	// Push notification processing
+	pushProcessor push.NotificationProcessor
+
+	// Hitless upgrade manager
+	hitlessManager     *hitless.HitlessManager
+	hitlessManagerLock sync.RWMutex
 }
 
 func (c *baseClient) clone() *baseClient {
-	clone := *c
-	return &clone
+	c.hitlessManagerLock.RLock()
+	hitlessManager := c.hitlessManager
+	c.hitlessManagerLock.RUnlock()
+
+	clone := &baseClient{
+		opt:            c.opt,
+		connPool:       c.connPool,
+		onClose:        c.onClose,
+		pushProcessor:  c.pushProcessor,
+		hitlessManager: hitlessManager,
+	}
+	return clone
 }
 
 func (c *baseClient) withTimeout(timeout time.Duration) *baseClient {
@@ -229,21 +251,6 @@ func (c *baseClient) String() string {
 	return fmt.Sprintf("Redis<%s db:%d>", c.getAddr(), c.opt.DB)
 }
 
-func (c *baseClient) newConn(ctx context.Context) (*pool.Conn, error) {
-	cn, err := c.connPool.NewConn(ctx)
-	if err != nil {
-		return nil, err
-	}
-
-	err = c.initConn(ctx, cn)
-	if err != nil {
-		_ = c.connPool.CloseConn(cn)
-		return nil, err
-	}
-
-	return cn, nil
-}
-
 func (c *baseClient) getConn(ctx context.Context) (*pool.Conn, error) {
 	if c.opt.Limiter != nil {
 		err := c.opt.Limiter.Allow()
@@ -269,7 +276,7 @@ func (c *baseClient) _getConn(ctx context.Context) (*pool.Conn, error) {
 		return nil, err
 	}
 
-	if cn.Inited {
+	if cn.IsInited() {
 		return cn, nil
 	}
 
@@ -351,12 +358,10 @@ func (c *baseClient) wrappedOnClose(newOnClose func() error) func() error {
 }
 
 func (c *baseClient) initConn(ctx context.Context, cn *pool.Conn) error {
-	if cn.Inited {
+	if !cn.Inited.CompareAndSwap(false, true) {
 		return nil
 	}
-
 	var err error
-	cn.Inited = true
 	connPool := pool.NewSingleConnPool(c.connPool, cn)
 	conn := newConn(c.opt, connPool, &c.hooksMixin)
 
@@ -425,6 +430,51 @@ func (c *baseClient) initConn(ctx context.Context, cn *pool.Conn) error {
 		return fmt.Errorf("failed to initialize connection options: %w", err)
 	}
 
+	// Enable maintenance notifications if hitless upgrades are configured
+	c.optLock.RLock()
+	hitlessEnabled := c.opt.HitlessUpgradeConfig != nil && c.opt.HitlessUpgradeConfig.Mode != hitless.MaintNotificationsDisabled
+	protocol := c.opt.Protocol
+	endpointType := c.opt.HitlessUpgradeConfig.EndpointType
+	c.optLock.RUnlock()
+	var hitlessHandshakeErr error
+	if hitlessEnabled && protocol == 3 {
+		hitlessHandshakeErr = conn.ClientMaintNotifications(
+			ctx,
+			true,
+			endpointType.String(),
+		).Err()
+		if hitlessHandshakeErr != nil {
+			if !isRedisError(hitlessHandshakeErr) {
+				// if not redis error, fail the connection
+				return hitlessHandshakeErr
+			}
+			c.optLock.Lock()
+			// handshake failed - check and modify config atomically
+			switch c.opt.HitlessUpgradeConfig.Mode {
+			case hitless.MaintNotificationsEnabled:
+				// enabled mode, fail the connection
+				c.optLock.Unlock()
+				return fmt.Errorf("failed to enable maintenance notifications: %w", hitlessHandshakeErr)
+			default: // will handle auto and any other
+				internal.Logger.Printf(ctx, "hitless: auto mode fallback: hitless upgrades disabled due to handshake error: %v", hitlessHandshakeErr)
+				c.opt.HitlessUpgradeConfig.Mode = hitless.MaintNotificationsDisabled
+				c.optLock.Unlock()
+				// auto mode, disable hitless upgrades and continue
+				if err := c.disableHitlessUpgrades(); err != nil {
+					// Log error but continue - auto mode should be resilient
+					internal.Logger.Printf(ctx, "hitless: failed to disable hitless upgrades in auto mode: %v", err)
+				}
+			}
+		} else {
+			// handshake was executed successfully
+			// to make sure that the handshake will be executed on other connections as well if it was successfully
+			// executed on this connection, we will force the handshake to be executed on all connections
+			c.optLock.Lock()
+			c.opt.HitlessUpgradeConfig.Mode = hitless.MaintNotificationsEnabled
+			c.optLock.Unlock()
+		}
+	}
+
 	if !c.opt.DisableIdentity && !c.opt.DisableIndentity {
 		libName := ""
 		libVer := Version()
@@ -441,6 +491,12 @@ func (c *baseClient) initConn(ctx context.Context, cn *pool.Conn) error {
 		}
 	}
 
+	cn.SetUsable(true)
+	cn.Inited.Store(true)
+
+	// Set the connection initialization function for potential reconnections
+	cn.SetInitConnFunc(c.createInitConnFunc())
+
 	if c.opt.OnConnect != nil {
 		return c.opt.OnConnect(ctx, conn)
 	}
@@ -456,6 +512,10 @@ func (c *baseClient) releaseConn(ctx context.Context, cn *pool.Conn, err error)
 	if isBadConn(err, false, c.opt.Addr) {
 		c.connPool.Remove(ctx, cn, err)
 	} else {
+		// process any pending push notifications before returning the connection to the pool
+		if err := c.processPushNotifications(ctx, cn); err != nil {
+			internal.Logger.Printf(ctx, "push: error processing pending notifications before releasing connection: %v", err)
+		}
 		c.connPool.Put(ctx, cn)
 	}
 }
@@ -497,16 +557,16 @@ func (c *baseClient) process(ctx context.Context, cmd Cmder) error {
 	return lastErr
 }
 
-func (c *baseClient) assertUnstableCommand(cmd Cmder) bool {
+func (c *baseClient) assertUnstableCommand(cmd Cmder) (bool, error) {
 	switch cmd.(type) {
 	case *AggregateCmd, *FTInfoCmd, *FTSpellCheckCmd, *FTSearchCmd, *FTSynDumpCmd:
 		if c.opt.UnstableResp3 {
-			return true
+			return true, nil
 		} else {
-			panic("RESP3 responses for this command are disabled because they may still change. Please set the flag UnstableResp3 .  See the [README](https://github.com/redis/go-redis/blob/master/README.md) and the release notes for guidance.")
+			return false, fmt.Errorf("RESP3 responses for this command are disabled because they may still change. Please set the flag UnstableResp3. See the README and the release notes for guidance")
 		}
 	default:
-		return false
+		return false, nil
 	}
 }
 
@@ -519,6 +579,11 @@ func (c *baseClient) _process(ctx context.Context, cmd Cmder, attempt int) (bool
 
 	retryTimeout := uint32(0)
 	if err := c.withConn(ctx, func(ctx context.Context, cn *pool.Conn) error {
+		// Process any pending push notifications before executing the command
+		if err := c.processPushNotifications(ctx, cn); err != nil {
+			internal.Logger.Printf(ctx, "push: error processing pending notifications before command: %v", err)
+		}
+
 		if err := cn.WithWriter(c.context(ctx), c.opt.WriteTimeout, func(wr *proto.Writer) error {
 			return writeCmd(wr, cmd)
 		}); err != nil {
@@ -527,10 +592,22 @@ func (c *baseClient) _process(ctx context.Context, cmd Cmder, attempt int) (bool
 		}
 		readReplyFunc := cmd.readReply
 		// Apply unstable RESP3 search module.
-		if c.opt.Protocol != 2 && c.assertUnstableCommand(cmd) {
+		if c.opt.Protocol != 2 {
+			useRawReply, err := c.assertUnstableCommand(cmd)
+			if err != nil {
+				return err
+			}
+			if useRawReply {
 				readReplyFunc = cmd.readRawReply
 			}
-		if err := cn.WithReader(c.context(ctx), c.cmdTimeout(cmd), readReplyFunc); err != nil {
+		}
+		if err := cn.WithReader(c.context(ctx), c.cmdTimeout(cmd), func(rd *proto.Reader) error {
+			// To be sure there are no buffered push notifications, we process them before reading the reply
+			if err := c.processPendingPushNotificationWithReader(ctx, cn, rd); err != nil {
+				internal.Logger.Printf(ctx, "push: error processing pending notifications before reading reply: %v", err)
+			}
+			return readReplyFunc(rd)
+		}); err != nil {
 			if cmd.readTimeout() == nil {
 				atomic.StoreUint32(&retryTimeout, 1)
 			} else {
@@ -573,20 +650,77 @@ func (c *baseClient) context(ctx context.Context) context.Context {
 	return context.Background()
 }
 
+// createInitConnFunc creates a connection initialization function that can be used for reconnections.
+func (c *baseClient) createInitConnFunc() func(context.Context, *pool.Conn) error {
+	return func(ctx context.Context, cn *pool.Conn) error {
+		return c.initConn(ctx, cn)
+	}
+}
+
+// enableHitlessUpgrades initializes the hitless upgrade manager and pool hook.
+// This function is called during client initialization.
+// will register push notification handlers for all hitless upgrade events.
+// will start background workers for handoff processing in the pool hook.
+func (c *baseClient) enableHitlessUpgrades() error {
+	// Create client adapter
+	clientAdapterInstance := newClientAdapter(c)
+
+	// Create hitless manager directly
+	manager, err := hitless.NewHitlessManager(clientAdapterInstance, c.connPool, c.opt.HitlessUpgradeConfig)
+	if err != nil {
+		return err
+	}
+	// Set the manager reference and initialize pool hook
+	c.hitlessManagerLock.Lock()
+	c.hitlessManager = manager
+	c.hitlessManagerLock.Unlock()
+
+	// Initialize pool hook (safe to call without lock since manager is now set)
+	manager.InitPoolHook(c.dialHook)
+	return nil
+}
+
+func (c *baseClient) disableHitlessUpgrades() error {
+	c.hitlessManagerLock.Lock()
+	defer c.hitlessManagerLock.Unlock()
+
+	// Close the hitless manager
+	if c.hitlessManager != nil {
+		// Closing the manager will also shutdown the pool hook
+		// and remove it from the pool
+		c.hitlessManager.Close()
+		c.hitlessManager = nil
+	}
+	return nil
+}
+
 // Close closes the client, releasing any open resources.
 //
 // It is rare to Close a Client, as the Client is meant to be
 // long-lived and shared between many goroutines.
 func (c *baseClient) Close() error {
 	var firstErr error
+
+	// Close hitless manager first
+	if err := c.disableHitlessUpgrades(); err != nil {
+		firstErr = err
+	}
+
 	if c.onClose != nil {
-		if err := c.onClose(); err != nil {
+		if err := c.onClose(); err != nil && firstErr == nil {
 			firstErr = err
 		}
 	}
+	if c.connPool != nil {
 		if err := c.connPool.Close(); err != nil && firstErr == nil {
 			firstErr = err
 		}
+	}
+	if c.pubSubPool != nil {
+		if err := c.pubSubPool.Close(); err != nil && firstErr == nil {
+			firstErr = err
+		}
+	}
 	return firstErr
 }
 
@@ -625,6 +759,10 @@ func (c *baseClient) generalProcessPipeline(
 		// Enable retries by default to retry dial errors returned by withConn.
 		canRetry := true
 		lastErr = c.withConn(ctx, func(ctx context.Context, cn *pool.Conn) error {
+			// Process any pending push notifications before executing the pipeline
+			if err := c.processPushNotifications(ctx, cn); err != nil {
+				internal.Logger.Printf(ctx, "push: error processing pending notifications before processing pipeline: %v", err)
+			}
 			var err error
 			canRetry, err = p(ctx, cn, cmds)
 			return err
@@ -640,6 +778,11 @@ func (c *baseClient) generalProcessPipeline(
 func (c *baseClient) pipelineProcessCmds(
 	ctx context.Context, cn *pool.Conn, cmds []Cmder,
 ) (bool, error) {
+	// Process any pending push notifications before executing the pipeline
+	if err := c.processPushNotifications(ctx, cn); err != nil {
+		internal.Logger.Printf(ctx, "push: error processing pending notifications before writing pipeline: %v", err)
+	}
+
 	if err := cn.WithWriter(c.context(ctx), c.opt.WriteTimeout, func(wr *proto.Writer) error {
 		return writeCmds(wr, cmds)
 	}); err != nil {
@@ -648,7 +791,8 @@ func (c *baseClient) pipelineProcessCmds(
 	}
 
 	if err := cn.WithReader(c.context(ctx), c.opt.ReadTimeout, func(rd *proto.Reader) error {
-		return pipelineReadCmds(rd, cmds)
+		// read all replies
+		return c.pipelineReadCmds(ctx, cn, rd, cmds)
 	}); err != nil {
 		return true, err
 	}
@@ -656,8 +800,12 @@ func (c *baseClient) pipelineProcessCmds(
 	return false, nil
 }
 
-func pipelineReadCmds(rd *proto.Reader, cmds []Cmder) error {
+func (c *baseClient) pipelineReadCmds(ctx context.Context, cn *pool.Conn, rd *proto.Reader, cmds []Cmder) error {
 	for i, cmd := range cmds {
+		// To be sure there are no buffered push notifications, we process them before reading the reply
+		if err := c.processPendingPushNotificationWithReader(ctx, cn, rd); err != nil {
+			internal.Logger.Printf(ctx, "push: error processing pending notifications before reading reply: %v", err)
+		}
 		err := cmd.readReply(rd)
 		cmd.SetErr(err)
 		if err != nil && !isRedisError(err) {
@@ -672,6 +820,11 @@ func pipelineReadCmds(rd *proto.Reader, cmds []Cmder) error {
 func (c *baseClient) txPipelineProcessCmds(
 	ctx context.Context, cn *pool.Conn, cmds []Cmder,
 ) (bool, error) {
+	// Process any pending push notifications before executing the transaction pipeline
+	if err := c.processPushNotifications(ctx, cn); err != nil {
+		internal.Logger.Printf(ctx, "push: error processing pending notifications before transaction: %v", err)
+	}
+
 	if err := cn.WithWriter(c.context(ctx), c.opt.WriteTimeout, func(wr *proto.Writer) error {
 		return writeCmds(wr, cmds)
 	}); err != nil {
@@ -684,12 +837,13 @@ func (c *baseClient) txPipelineProcessCmds(
 		// Trim multi and exec.
 		trimmedCmds := cmds[1 : len(cmds)-1]
 
-		if err := txPipelineReadQueued(rd, statusCmd, trimmedCmds); err != nil {
+		if err := c.txPipelineReadQueued(ctx, cn, rd, statusCmd, trimmedCmds); err != nil {
 			setCmdsErr(cmds, err)
 			return err
 		}
 
-		return pipelineReadCmds(rd, trimmedCmds)
+		// Read replies.
+		return c.pipelineReadCmds(ctx, cn, rd, trimmedCmds)
 	}); err != nil {
 		return false, err
 	}
@@ -697,7 +851,13 @@ func (c *baseClient) txPipelineProcessCmds(
 	return false, nil
 }
 
-func txPipelineReadQueued(rd *proto.Reader, statusCmd *StatusCmd, cmds []Cmder) error {
+// txPipelineReadQueued reads queued replies from the Redis server.
+// It returns an error if the server returns an error or if the number of replies does not match the number of commands.
+func (c *baseClient) txPipelineReadQueued(ctx context.Context, cn *pool.Conn, rd *proto.Reader, statusCmd *StatusCmd, cmds []Cmder) error {
+	// To be sure there are no buffered push notifications, we process them before reading the reply
+	if err := c.processPendingPushNotificationWithReader(ctx, cn, rd); err != nil {
+		internal.Logger.Printf(ctx, "push: error processing pending notifications before reading reply: %v", err)
+	}
 	// Parse +OK.
 	if err := statusCmd.readReply(rd); err != nil {
 		return err
@@ -705,6 +865,10 @@ func txPipelineReadQueued(rd *proto.Reader, statusCmd *StatusCmd, cmds []Cmder)
 
 	// Parse +QUEUED.
   for _, cmd := range cmds {
+		// To be sure there are no buffered push notifications, we process them before reading the reply
+		if err := c.processPendingPushNotificationWithReader(ctx, cn, rd); err != nil {
+			internal.Logger.Printf(ctx, "push: error processing pending notifications before reading reply: %v", err)
+		}
 		if err := statusCmd.readReply(rd); err != nil {
 			cmd.SetErr(err)
 			if !isRedisError(err) {
@@ -713,6 +877,10 @@ func txPipelineReadQueued(rd *proto.Reader, statusCmd *StatusCmd, cmds []Cmder)
 		}
 	}
 
+	// To be sure there are no buffered push notifications, we process them before reading the reply
+	if err := c.processPendingPushNotificationWithReader(ctx, cn, rd); err != nil {
+		internal.Logger.Printf(ctx, "push: error processing pending notifications before reading reply: %v", err)
+	}
 	// Parse number of replies.
 	line, err := rd.ReadLine()
 	if err != nil {
@@ -746,15 +914,56 @@ func NewClient(opt *Options) *Client {
 	if opt == nil {
 		panic("redis: NewClient nil options")
 	}
+	// clone to not share options with the caller
+	opt = opt.clone()
 	opt.init()
 
+	// Push notifications are always enabled for RESP3 (cannot be disabled)
+
 	c := Client{
 		baseClient: &baseClient{
 			opt: opt,
 		},
 	}
 	c.init()
-	c.connPool = newConnPool(opt, c.dialHook)
+
+	// Initialize push notification processor using shared helper
+	// Use void processor for RESP2 connections (push notifications not available)
+	c.pushProcessor = initializePushProcessor(opt)
+	// set opt push processor for child clients
+	c.opt.PushNotificationProcessor = c.pushProcessor
+
+	// Create connection pools
+	var err error
+	c.connPool, err = newConnPool(opt, c.dialHook)
+	if err != nil {
+		panic(fmt.Errorf("redis: failed to create connection pool: %w", err))
+	}
+	c.pubSubPool, err = newPubSubPool(opt, c.dialHook)
+	if err != nil {
+		panic(fmt.Errorf("redis: failed to create pubsub pool: %w", err))
+	}
+
+	// Initialize hitless upgrades first if enabled and protocol is RESP3
+	if opt.HitlessUpgradeConfig != nil && opt.HitlessUpgradeConfig.Mode != hitless.MaintNotificationsDisabled && opt.Protocol == 3 {
+		err := c.enableHitlessUpgrades()
+		if err != nil {
+			internal.Logger.Printf(context.Background(), "hitless: failed to initialize hitless upgrades: %v", err)
+			if opt.HitlessUpgradeConfig.Mode == hitless.MaintNotificationsEnabled {
+				/*
+					Design decision: panic here to fail fast if hitless upgrades cannot be enabled when explicitly requested.
+					We choose to panic instead of returning an error to avoid breaking the existing client API, which does not expect
+					an error from NewClient. This ensures that misconfiguration or critical initialization failures are surfaced
+					immediately, rather than allowing the client to continue in a partially initialized or inconsistent state.
+					Clients relying on hitless upgrades should be aware that initialization errors will cause a panic, and should
+					handle this accordingly (e.g., via recover or by validating configuration before calling NewClient).
+					This approach is only used when HitlessUpgradeConfig.Mode is MaintNotificationsEnabled, indicating that hitless
+					upgrades are required for correct operation. In other modes, initialization failures are logged but do not panic.
+				*/
+				panic(fmt.Errorf("failed to enable hitless upgrades: %w", err))
+			}
+		}
+	}
 
 	return &c
 }
@@ -791,11 +1000,51 @@ func (c *Client) Options() *Options {
 	return c.opt
 }
 
+// GetHitlessManager returns the hitless manager instance for monitoring and control.
+// Returns nil if hitless upgrades are not enabled.
+func (c *Client) GetHitlessManager() *hitless.HitlessManager {
+	c.hitlessManagerLock.RLock()
+	defer c.hitlessManagerLock.RUnlock()
+	return c.hitlessManager
+}
+
+// initializePushProcessor initializes the push notification processor for any client type.
+// This is a shared helper to avoid duplication across NewClient, NewFailoverClient, and NewSentinelClient.
+func initializePushProcessor(opt *Options) push.NotificationProcessor {
+	// Always use custom processor if provided
+	if opt.PushNotificationProcessor != nil {
+		return opt.PushNotificationProcessor
+	}
+
+	// Push notifications are always enabled for RESP3, disabled for RESP2
+	if opt.Protocol == 3 {
+		// Create default processor for RESP3 connections
+		return NewPushNotificationProcessor()
+	}
+
+	// Create void processor for RESP2 connections (push notifications not available)
+	return NewVoidPushNotificationProcessor()
+}
+
+// RegisterPushNotificationHandler registers a handler for a specific push notification name.
+// Returns an error if a handler is already registered for this push notification name.
+// If protected is true, the handler cannot be unregistered.
+func (c *Client) RegisterPushNotificationHandler(pushNotificationName string, handler push.NotificationHandler, protected bool) error {
+	return c.pushProcessor.RegisterHandler(pushNotificationName, handler, protected)
+}
+
+// GetPushNotificationHandler returns the handler for a specific push notification name.
+// Returns nil if no handler is registered for the given name.
+func (c *Client) GetPushNotificationHandler(pushNotificationName string) push.NotificationHandler {
+	return c.pushProcessor.GetHandler(pushNotificationName)
+}
+
 type PoolStats pool.Stats
 
 // PoolStats returns connection pool stats.
 func (c *Client) PoolStats() *PoolStats {
 	stats := c.connPool.Stats()
+	stats.PubSubStats = *(c.pubSubPool.Stats())
 	return (*PoolStats)(stats)
 }
 
@@ -830,13 +1079,31 @@ func (c *Client) TxPipeline() Pipeliner {
 func (c *Client) pubSub() *PubSub {
 	pubsub := &PubSub{
 		opt: c.opt,
-
-		newConn: func(ctx context.Context, channels []string) (*pool.Conn, error) {
-			return c.newConn(ctx)
+		newConn: func(ctx context.Context, addr string, channels []string) (*pool.Conn, error) {
+			cn, err := c.pubSubPool.NewConn(ctx, c.opt.Network, addr, channels)
+			if err != nil {
+				return nil, err
+			}
+			// will return nil if already initialized
+			err = c.initConn(ctx, cn)
+			if err != nil {
+				_ = cn.Close()
+				return nil, err
+			}
+			// Track connection in PubSubPool
+			c.pubSubPool.TrackConn(cn)
+			return cn, nil
 		},
-		closeConn: c.connPool.CloseConn,
+		closeConn: func(cn *pool.Conn) error {
+			// Untrack connection from PubSubPool
+			c.pubSubPool.UntrackConn(cn)
+			_ = cn.Close()
+			return nil
+		},
+		pushProcessor: c.pushProcessor,
 	}
 	pubsub.init()
+
 	return pubsub
 }
 
@@ -920,6 +1187,10 @@ func newConn(opt *Options, connPool pool.Pooler, parentHooks *hooksMixin) *Conn
 		c.hooksMixin = parentHooks.clone()
 	}
 
+	// Initialize push notification processor using shared helper
+	// Use void processor for RESP2 connections (push notifications not available)
+	c.pushProcessor = initializePushProcessor(opt)
+
 	c.cmdable = c.Process
 	c.statefulCmdable = c.Process
 	c.initHooks(hooks{
@@ -938,6 +1209,13 @@ func (c *Conn) Process(ctx context.Context, cmd Cmder) error {
 	return err
 }
 
+// RegisterPushNotificationHandler registers a handler for a specific push notification name.
+// Returns an error if a handler is already registered for this push notification name.
+// If protected is true, the handler cannot be unregistered.
+func (c *Conn) RegisterPushNotificationHandler(pushNotificationName string, handler push.NotificationHandler, protected bool) error {
+	return c.pushProcessor.RegisterHandler(pushNotificationName, handler, protected)
+}
+
 func (c *Conn) Pipelined(ctx context.Context, fn func(Pipeliner) error) ([]Cmder, error) {
 	return c.Pipeline().Pipelined(ctx, fn)
 }
@@ -965,3 +1243,50 @@ func (c *Conn) TxPipeline() Pipeliner {
 	pipe.init()
 	return &pipe
 }
+
+// processPushNotifications processes all pending push notifications on a connection
+// This ensures that cluster topology changes are handled immediately before the connection is used
+// This method should be called by the client before using WithReader for command execution
+func (c *baseClient) processPushNotifications(ctx context.Context, cn *pool.Conn) error {
+	// Only process push notifications for RESP3 connections with a processor
+	// Also check if there is any data to read before processing
+	// Which is an optimization on UNIX systems where MaybeHasData is a syscall
+	// On Windows, MaybeHasData always returns true, so this check is a no-op
+	if c.opt.Protocol != 3 || c.pushProcessor == nil || !cn.MaybeHasData() {
+		return nil
+	}
+
+	// Use WithReader to access the reader and process push notifications
+	// This is critical for hitless upgrades to work properly
+	// NOTE: almost no timeouts are set for this read, so it should not block
+	// longer than necessary, 10us should be plenty of time to read if there are any push notifications
+	// on the socket.
+	return cn.WithReader(ctx, 10*time.Microsecond, func(rd *proto.Reader) error {
+		// Create handler context with client, connection pool, and connection information
+		handlerCtx := c.pushNotificationHandlerContext(cn)
+		return c.pushProcessor.ProcessPendingNotifications(ctx, handlerCtx, rd)
+	})
+}
+
+// processPendingPushNotificationWithReader processes all pending push notifications on a connection
+// This method should be called by the client in WithReader before reading the reply
+func (c *baseClient) processPendingPushNotificationWithReader(ctx context.Context, cn *pool.Conn, rd *proto.Reader) error {
+	// if we have the reader, we don't need to check for data on the socket, we are waiting
+	// for either a reply or a push notification, so we can block until we get a reply or reach the timeout
+	if c.opt.Protocol != 3 || c.pushProcessor == nil {
+		return nil
+	}
+
+	// Create handler context with client, connection pool, and connection information
+	handlerCtx := c.pushNotificationHandlerContext(cn)
+	return c.pushProcessor.ProcessPendingNotifications(ctx, handlerCtx, rd)
+}
+
+// pushNotificationHandlerContext creates a handler context for push notification processing
+func (c *baseClient) pushNotificationHandlerContext(cn *pool.Conn) push.NotificationHandlerContext {
+	return push.NotificationHandlerContext{
+		Client:   c,
+		ConnPool: c.connPool,
+		Conn:     cn, // Wrap in adapter for easier interface access
+	}
+}

redis_test.go

@@ -12,7 +12,6 @@ import (
 
 	. "github.com/bsm/ginkgo/v2"
 	. "github.com/bsm/gomega"
-
 	"github.com/redis/go-redis/v9"
 	"github.com/redis/go-redis/v9/auth"
 )

search_test.go

@@ -3407,14 +3407,16 @@ var _ = Describe("RediSearch commands Resp 3", Label("search"), func() {
 		Expect(rawValResults[0]).To(Or(BeEquivalentTo(results[0]), BeEquivalentTo(results[1])))
 		Expect(rawValResults[1]).To(Or(BeEquivalentTo(results[0]), BeEquivalentTo(results[1])))
 
-		// Test with UnstableResp3 false
-		Expect(func() {
+		// Test with UnstableResp3 false - should return error instead of panic
 		options = &redis.FTAggregateOptions{Apply: []redis.FTAggregateApply{{Field: "@CreatedDateTimeUTC * 10", As: "CreatedDateTimeUTC"}}}
-			rawRes, _ := client2.FTAggregateWithArgs(ctx, "idx1", "*", options).RawResult()
-			rawVal = client2.FTAggregateWithArgs(ctx, "idx1", "*", options).RawVal()
+		rawRes, err := client2.FTAggregateWithArgs(ctx, "idx1", "*", options).RawResult()
+		Expect(err).To(HaveOccurred())
+		Expect(err.Error()).To(ContainSubstring("RESP3 responses for this command are disabled"))
 		Expect(rawRes).To(BeNil())
+
+		rawVal = client2.FTAggregateWithArgs(ctx, "idx1", "*", options).RawVal()
+		Expect(client2.FTAggregateWithArgs(ctx, "idx1", "*", options).Err()).To(HaveOccurred())
 		Expect(rawVal).To(BeNil())
-		}).Should(Panic())
 
 	})
 
@@ -3435,13 +3437,15 @@ var _ = Describe("RediSearch commands Resp 3", Label("search"), func() {
 		flags = attributes[0].(map[interface{}]interface{})["flags"].([]interface{})
 		Expect(flags).To(ConsistOf("SORTABLE", "NOSTEM"))
 
-		// Test with UnstableResp3 false
-		Expect(func() {
-			rawResInfo, _ := client2.FTInfo(ctx, "idx1").RawResult()
-			rawValInfo := client2.FTInfo(ctx, "idx1").RawVal()
+		// Test with UnstableResp3 false - should return error instead of panic
+		rawResInfo, err := client2.FTInfo(ctx, "idx1").RawResult()
+		Expect(err).To(HaveOccurred())
+		Expect(err.Error()).To(ContainSubstring("RESP3 responses for this command are disabled"))
 		Expect(rawResInfo).To(BeNil())
+
+		rawValInfo := client2.FTInfo(ctx, "idx1").RawVal()
+		Expect(client2.FTInfo(ctx, "idx1").Err()).To(HaveOccurred())
 		Expect(rawValInfo).To(BeNil())
-		}).Should(Panic())
 	})
 
 	It("should handle FTSpellCheck with Unstable RESP3 Search Module and without stability", Label("search", "ftcreate", "ftspellcheck"), func() {
@@ -3462,13 +3466,15 @@ var _ = Describe("RediSearch commands Resp 3", Label("search"), func() {
 		results := resSpellCheck.(map[interface{}]interface{})["results"].(map[interface{}]interface{})
 		Expect(results["impornant"].([]interface{})[0].(map[interface{}]interface{})["important"]).To(BeEquivalentTo(0.5))
 
-		// Test with UnstableResp3 false
-		Expect(func() {
-			rawResSpellCheck, _ := client2.FTSpellCheck(ctx, "idx1", "impornant").RawResult()
-			rawValSpellCheck := client2.FTSpellCheck(ctx, "idx1", "impornant").RawVal()
+		// Test with UnstableResp3 false - should return error instead of panic
+		rawResSpellCheck, err := client2.FTSpellCheck(ctx, "idx1", "impornant").RawResult()
+		Expect(err).To(HaveOccurred())
+		Expect(err.Error()).To(ContainSubstring("RESP3 responses for this command are disabled"))
 		Expect(rawResSpellCheck).To(BeNil())
+
+		rawValSpellCheck := client2.FTSpellCheck(ctx, "idx1", "impornant").RawVal()
+		Expect(client2.FTSpellCheck(ctx, "idx1", "impornant").Err()).To(HaveOccurred())
 		Expect(rawValSpellCheck).To(BeNil())
-		}).Should(Panic())
 	})
 
 	It("should handle FTSearch with Unstable RESP3 Search Module and without stability", Label("search", "ftcreate", "ftsearch"), func() {
@@ -3489,13 +3495,15 @@ var _ = Describe("RediSearch commands Resp 3", Label("search"), func() {
 		totalResults2 := res2.(map[interface{}]interface{})["total_results"]
 		Expect(totalResults2).To(BeEquivalentTo(int64(1)))
 
-		// Test with UnstableResp3 false
-		Expect(func() {
-			rawRes2, _ := client2.FTSearchWithArgs(ctx, "txt", "foo bar hello world", &redis.FTSearchOptions{NoContent: true}).RawResult()
-			rawVal2 := client2.FTSearchWithArgs(ctx, "txt", "foo bar hello world", &redis.FTSearchOptions{NoContent: true}).RawVal()
+		// Test with UnstableResp3 false - should return error instead of panic
+		rawRes2, err := client2.FTSearchWithArgs(ctx, "txt", "foo bar hello world", &redis.FTSearchOptions{NoContent: true}).RawResult()
+		Expect(err).To(HaveOccurred())
+		Expect(err.Error()).To(ContainSubstring("RESP3 responses for this command are disabled"))
 		Expect(rawRes2).To(BeNil())
+
+		rawVal2 := client2.FTSearchWithArgs(ctx, "txt", "foo bar hello world", &redis.FTSearchOptions{NoContent: true}).RawVal()
+		Expect(client2.FTSearchWithArgs(ctx, "txt", "foo bar hello world", &redis.FTSearchOptions{NoContent: true}).Err()).To(HaveOccurred())
 		Expect(rawVal2).To(BeNil())
-		}).Should(Panic())
 	})
 	It("should handle FTSynDump with Unstable RESP3 Search Module and without stability", Label("search", "ftsyndump"), func() {
 		text1 := &redis.FieldSchema{FieldName: "title", FieldType: redis.SearchFieldTypeText}
@@ -3523,13 +3531,15 @@ var _ = Describe("RediSearch commands Resp 3", Label("search"), func() {
 		Expect(valSynDump).To(BeEquivalentTo(resSynDump))
 		Expect(resSynDump.(map[interface{}]interface{})["baby"]).To(BeEquivalentTo([]interface{}{"id1"}))
 
-		// Test with UnstableResp3 false
-		Expect(func() {
-			rawResSynDump, _ := client2.FTSynDump(ctx, "idx1").RawResult()
-			rawValSynDump := client2.FTSynDump(ctx, "idx1").RawVal()
+		// Test with UnstableResp3 false - should return error instead of panic
+		rawResSynDump, err := client2.FTSynDump(ctx, "idx1").RawResult()
+		Expect(err).To(HaveOccurred())
+		Expect(err.Error()).To(ContainSubstring("RESP3 responses for this command are disabled"))
 		Expect(rawResSynDump).To(BeNil())
+
+		rawValSynDump := client2.FTSynDump(ctx, "idx1").RawVal()
+		Expect(client2.FTSynDump(ctx, "idx1").Err()).To(HaveOccurred())
 		Expect(rawValSynDump).To(BeNil())
-		}).Should(Panic())
 	})
 
 	It("should test not affected Resp 3 Search method - FTExplain", Label("search", "ftexplain"), func() {

sentinel.go

@@ -17,6 +17,7 @@ import (
 	"github.com/redis/go-redis/v9/internal/pool"
 	"github.com/redis/go-redis/v9/internal/rand"
 	"github.com/redis/go-redis/v9/internal/util"
+	"github.com/redis/go-redis/v9/push"
 )
 
 //------------------------------------------------------------------------------
@@ -62,6 +63,8 @@ type FailoverOptions struct {
 	Protocol int
 	Username string
 	Password string
+
+	// Push notifications are always enabled for RESP3 connections
 	// CredentialsProvider allows the username and password to be updated
 	// before reconnecting. It should return the current username and password.
 	CredentialsProvider func() (username string, password string)
@@ -136,6 +139,14 @@ type FailoverOptions struct {
 	FailingTimeoutSeconds int
 
 	UnstableResp3 bool
+
+	// Hitless is not supported for FailoverClients at the moment
+	// HitlessUpgradeConfig provides custom configuration for hitless upgrades.
+	// When HitlessUpgradeConfig.Mode is not "disabled", the client will handle
+	// upgrade notifications gracefully and manage connection/pool state transitions
+	// seamlessly. Requires Protocol: 3 (RESP3) for push notifications.
+	// If nil, hitless upgrades are disabled.
+	//HitlessUpgradeConfig *HitlessUpgradeConfig
 }
 
 func (opt *FailoverOptions) clientOptions() *Options {
@@ -454,8 +465,6 @@ func NewFailoverClient(failoverOpt *FailoverOptions) *Client {
 	opt.Dialer = masterReplicaDialer(failover)
 	opt.init()
 
-	var connPool *pool.ConnPool
-
 	rdb := &Client{
 		baseClient: &baseClient{
 			opt: opt,
@@ -463,16 +472,30 @@ func NewFailoverClient(failoverOpt *FailoverOptions) *Client {
 	}
 	rdb.init()
 
-	connPool = newConnPool(opt, rdb.dialHook)
-	rdb.connPool = connPool
+	// Initialize push notification processor using shared helper
+	// Use void processor by default for RESP2 connections
+	rdb.pushProcessor = initializePushProcessor(opt)
+
+	var err error
+	rdb.connPool, err = newConnPool(opt, rdb.dialHook)
+	if err != nil {
+		panic(fmt.Errorf("redis: failed to create connection pool: %w", err))
+	}
+	rdb.pubSubPool, err = newPubSubPool(opt, rdb.dialHook)
+	if err != nil {
+		panic(fmt.Errorf("redis: failed to create pubsub pool: %w", err))
+	}
+
 	rdb.onClose = rdb.wrappedOnClose(failover.Close)
 
 	failover.mu.Lock()
 	failover.onFailover = func(ctx context.Context, addr string) {
+		if connPool, ok := rdb.connPool.(*pool.ConnPool); ok {
 			_ = connPool.Filter(func(cn *pool.Conn) bool {
 				return cn.RemoteAddr().String() != addr
 			})
 		}
+	}
 	failover.mu.Unlock()
 
 	return rdb
@@ -529,15 +552,40 @@ func NewSentinelClient(opt *Options) *SentinelClient {
 		},
 	}
 
+	// Initialize push notification processor using shared helper
+	// Use void processor for Sentinel clients
+	c.pushProcessor = NewVoidPushNotificationProcessor()
+
 	c.initHooks(hooks{
 		dial:    c.baseClient.dial,
 		process: c.baseClient.process,
 	})
-	c.connPool = newConnPool(opt, c.dialHook)
+	var err error
+	c.connPool, err = newConnPool(opt, c.dialHook)
+	if err != nil {
+		panic(fmt.Errorf("redis: failed to create connection pool: %w", err))
+	}
+	c.pubSubPool, err = newPubSubPool(opt, c.dialHook)
+	if err != nil {
+		panic(fmt.Errorf("redis: failed to create pubsub pool: %w", err))
+	}
 
 	return c
 }
 
+// GetPushNotificationHandler returns the handler for a specific push notification name.
+// Returns nil if no handler is registered for the given name.
+func (c *SentinelClient) GetPushNotificationHandler(pushNotificationName string) push.NotificationHandler {
+	return c.pushProcessor.GetHandler(pushNotificationName)
+}
+
+// RegisterPushNotificationHandler registers a handler for a specific push notification name.
+// Returns an error if a handler is already registered for this push notification name.
+// If protected is true, the handler cannot be unregistered.
+func (c *SentinelClient) RegisterPushNotificationHandler(pushNotificationName string, handler push.NotificationHandler, protected bool) error {
+	return c.pushProcessor.RegisterHandler(pushNotificationName, handler, protected)
+}
+
 func (c *SentinelClient) Process(ctx context.Context, cmd Cmder) error {
 	err := c.processHook(ctx, cmd)
 	cmd.SetErr(err)
@@ -547,13 +595,31 @@ func (c *SentinelClient) Process(ctx context.Context, cmd Cmder) error {
 func (c *SentinelClient) pubSub() *PubSub {
 	pubsub := &PubSub{
 		opt: c.opt,
-
-		newConn: func(ctx context.Context, channels []string) (*pool.Conn, error) {
-			return c.newConn(ctx)
+		newConn: func(ctx context.Context, addr string, channels []string) (*pool.Conn, error) {
+			cn, err := c.pubSubPool.NewConn(ctx, c.opt.Network, addr, channels)
+			if err != nil {
+				return nil, err
+			}
+			// will return nil if already initialized
+			err = c.initConn(ctx, cn)
+			if err != nil {
+				_ = cn.Close()
+				return nil, err
+			}
+			// Track connection in PubSubPool
+			c.pubSubPool.TrackConn(cn)
+			return cn, nil
 		},
-		closeConn: c.connPool.CloseConn,
+		closeConn: func(cn *pool.Conn) error {
+			// Untrack connection from PubSubPool
+			c.pubSubPool.UntrackConn(cn)
+			_ = cn.Close()
+			return nil
+		},
+		pushProcessor: c.pushProcessor,
 	}
 	pubsub.init()
+
 	return pubsub
 }
 

tx.go

@@ -24,9 +24,10 @@ type Tx struct {
 func (c *Client) newTx() *Tx {
 	tx := Tx{
 		baseClient: baseClient{
-			opt:        c.opt,
+			opt:           c.opt.clone(), // Clone options to avoid sharing mutable state between transaction and parent client
 			connPool:      pool.NewStickyConnPool(c.connPool),
 			hooksMixin:    c.hooksMixin.clone(),
+			pushProcessor: c.pushProcessor, // Copy push processor from parent client
 		},
 	}
 	tx.init()

universal.go

@@ -122,6 +122,9 @@ type UniversalOptions struct {
 
 	// IsClusterMode can be used when only one Addrs is provided (e.g. Elasticache supports setting up cluster mode with configuration endpoint).
 	IsClusterMode bool
+
+	// HitlessUpgradeConfig provides configuration for hitless upgrades.
+	HitlessUpgradeConfig *HitlessUpgradeConfig
 }
 
 // Cluster returns cluster options created from the universal options.
@@ -177,6 +180,7 @@ func (o *UniversalOptions) Cluster() *ClusterOptions {
 		IdentitySuffix:        o.IdentitySuffix,
 		FailingTimeoutSeconds: o.FailingTimeoutSeconds,
 		UnstableResp3:         o.UnstableResp3,
+		HitlessUpgradeConfig:  o.HitlessUpgradeConfig,
 	}
 }
 
@@ -237,6 +241,7 @@ func (o *UniversalOptions) Failover() *FailoverOptions {
 		DisableIndentity: o.DisableIndentity,
 		IdentitySuffix:   o.IdentitySuffix,
 		UnstableResp3:    o.UnstableResp3,
+		// Note: HitlessUpgradeConfig not supported for FailoverOptions
 	}
 }
 
@@ -288,6 +293,7 @@ func (o *UniversalOptions) Simple() *Options {
 		DisableIndentity:     o.DisableIndentity,
 		IdentitySuffix:       o.IdentitySuffix,
 		UnstableResp3:        o.UnstableResp3,
+		HitlessUpgradeConfig: o.HitlessUpgradeConfig,
 	}
 }