redis/node-redis
1
0
Fork 0
mirror of https://github.com/redis/node-redis.git synced 2025-08-07 13:22:56 +03:00
Code Activity

V4 (#1624)

Browse Source 
* init v4

* add .gitignore to benchmark

* spawn redis-servers for tests,
add some tests,
fix client auth on connect

* add tests coverage report

* add tests workflow, replace nyc text reporter with text-summary

* run tests with node 16.x & redis 6.x only (for now)

* add socket events on client,
stop reconnectiong when manually calling disconnect,
remove abort signal listener when a command is written on the socket

* add isOpen boolean getter on client, add maxLength option to command queue, add test for client.multi

* move to use CommonJS

* add MULTI and EXEC commands to when executing multi command, make client.multi return type innerit the module commands, clean some tests, exclute spec files from coverage report

* missing file from commit 61edd4f1b5

* exclude spec files from coverage report

* add support for options in a command function (.get, .set, ...), add support for the SELECT command, implement a couple of commands, fix client socket reconnection strategy, add support for using replicas (RO) in cluster, and more..

* fix client.blPop test

* use which to find redis-server path

* change command options to work with Symbol rather then WeakSet

* implement more commands

* Add support for lua scripts in client & muilti, fix client socket initiator, implement simple cluster nodes discovery strategy

* replace `callbackify` with `legacyMode`

* add the SCAN command and client.scanIterator

* rename scanIterator

* init benchmark workflow

* fix benchmark workflow

* fix benchmark workflow

* fix benchmark workflow

* push coverage report to Coveralls

* fix Coveralls

* generator lcov (for Coveralls)

* fix .nycrc.json

* PubSub

* add support for all set commands (including sScanIterator)

* support pipeline

* fix KEEPTTL in SET

* remove console.log

* add HyperLogLog commands

* update README.md (thanks to @guyroyse)

* add support for most of the "keys commands"

* fix EXPIREAT.spec.ts

* add support for date in both EXPIREAT & EXPIRE

* add tests

* better cluster nodes discorvery strategy after MOVED error, add PubSub test

* fix PubSub UNSUBSCRIBE/PUNSUBSCRIBE without channel and/or listener

* fix PubSub

* add release-it to dev dependencies

* Release 4.0.0-next.0

* fix .npmignore

* Release 4.0.0-next.1

* fix links in README.md

* fix .npmignore

* Release 4.0.0-next.2

* add support for all sorted set commands

* add support for most stream commands

* add missing file from commit 53de279afe

* lots of todo commends

* make PubSub test more stable

* clean ZPOPMAX

* add support for lua scripts and modules in cluster, spawn cluster for tests, add some cluster tests, fix pubsub listener arguments

* GET.spec.ts

* add support for List commands, fix some Sorted Set commands, add some cluster commands, spawn cluster for testing, add support for command options in cluster, and more

* add missing file from commit faab94fab2

* clean ZRANK and ZREVRANK

* add XREAD and XREADGROUP commands

* remove unused files

* implement a couple of more commands, make cluster random iterator be per node (instead of per slot)

* Release 4.0.0-next.3

* app spec files to npmignore

* fix some code analyzers (LGTM, deepsource, codeclimate) issues

* fix CLUSTER_NODES, add some tests

* add HSCAN, clean some commands, add tests for generic transformers

* add missing files from 0feb35a1fb

* update README.md (thanks to @guyroyse)

* handle ASK errors, add some commands and tests

* Release 4.0.0-next.4

* replace "modern" with "v4"

* remove unused imports

* add all ACL subcommands, all MODULE subcommands, and some other commands

* remove 2 unused imports

* fix BITFIELD command

* fix XTRIM spec file

* clean code

* fix package.json types field

* better modules support, fix some bugs in legacy mode, add some tests

* remove unused function

* add test for hScanIterator

* change node mimimum version to 12 (latest LTS)

* update tsconfig.json to support node 12, run tests on Redis 5 & 6 and on all node live versions

* remove future node releases :P

* remove "lib" from ts compiler options

* Update tsconfig.json

* fix build

* run some tests only on supported redis versions, use coveralls parallel mode

* fix tests

* Do not use "timers/promises", fix "isRedisVersionGreaterThan"

* skip AbortController tests when not available

* use 'fs'.promises instead of 'fs/promises'

* add some missing commands

* run GETDEL tests only if the redis version is greater than 6.2

* implement some GEO commands, improve scan generic transformer, expose RPUSHX

* fix GEOSEARCH & GEOSEARCHSTORE

* use socket.setNoDelay and queueMicrotask to improve latency

* commands-queue.ts: String length / byte length counting issue (#1630)

* Update commands-queue.ts

Hopefully fixing #1628

* Reverted 2fa5ea6, and implemented test for byte length check

* Changed back to Buffer.byteLength, due to issue author input. Updated test to look for 4 bytes.

* Fixed. There were two places that length was calculated.

* Removed redundant string assignment

* add 2 bytes test as well

Co-authored-by: Leibale Eidelman <leibale1998@gmail.com>

* fix scripts in multi

* do not hide bugs in redis

* fix for e7bf09644b

* remove unused import

* implement WATCH command, fix ZRANGESTORE & GEOSEARCHSTORE tests

* update README.md

Co-authored-by: @GuyRoyse

* use typedoc to auto generate documentation

* run "npm install" before "npm run documentation"

* clean documentation workflow

* fix WATCH spec file

* increase "CLUSTER_NODE_TIMEOUT" to 5000ms to avoid "CLUSTERDOWN" errors in tests

* pull cluster state every 100 ms

* await meetPromises before pulling the cluster state

* enhance the way commanders (client/multi/cluster) get extended with modules and scripts

* add test for socket retry strategy

* implement more commands

* set GETEX minimum version to 6.2

* remove unused imports

* add support for multi in cluster

* upgrade dependencies

* Release 4.0.0-next.5

* remove unused imports

* improve benchmarking

* use the same Multi with duplicated clients

* exclude some files from the documentation, add some exports, clean code

* fix #1636 - handle null in multi.exec

* remove unused import

* add supoprt for tuples in HSET

* add FIRST_KEY_INDEX to HSET

* add a bunch of missing commands, fix MSET and HELLO, add some tests

* add FIRST_KEY_INDEX to MSET and MSETNX

* upgrade actions

* fix coverallsapp/github-action version

* Update documentation.yml

* Update documentation.yml

* clean code

* remove unused imports

* use "npm ci" instead of "npm install"

* fix `self` binding on client modules, use connection pool for `duplicateConnection`

* add client.executeIsolated, rename "duplicateConnection" to "isolated", update README.md (thanks to @GuyRoyse and @SimonPrickett)

* update README (thanks to @GuyRoyse), add some tests

* try to fix "cluster is down" errors in tests

* try to fix "cluster is down" errors in tests

* upgrade dependencies

* update package-lock

* Release 4.0.0-next.6

* fix #1636 - fix WatchError

* fix for f1bf0beebf - remove .only from multi tests

* Release 4.0.0-next.7

* update README and other markdown files

Co-authored-by: @GuyRoyse & @SimonPrickett

* Doc updates. (#1640)

* update docs, upgrade dependencies

* fix README

* Release 4.0.0-rc.0

* Update README.md

* update docs, add `connectTimeout` options, fix tls

Co-authored-by: Guy Royse <guy@guyroyse.com>

* npm update, "fix" some tests, clean code

* fix AssertionError import

* fix #1642 - fix XREAD, XREADGROUP and XTRIM

* fix #1644 - add the QUIT command

* add socket.noDelay and socket.keepAlive configurations

* Update README.md (#1645)

* Update README.md

Fixed issue with how connection string was specified.
Now you can have user@host without having to specify a password, which just makes more sense

* Update client-configuration.md as well

Co-authored-by: Leibale Eidelman <leibale1998@gmail.com>

* update socket.reconnectStrategy description

* fix borken link in v3-to-v4.md

* increase test coverage, fix bug in cluster redirection strategy, implement CLIENT_ID, remove unused EXEC command

Co-authored-by: Nova <novaw@warrenservices.co.uk>
Co-authored-by: Simon Prickett <simon@crudworks.org>
Co-authored-by: Guy Royse <guy@guyroyse.com>
This commit is contained in:
Leibale Eidelman
2021-09-02 10:04:48 -04:00
committed by GitHub
parent 4f85030e42
commit 4e6d018d77
661 changed files with 28847 additions and 14559 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
docs/FAQ.md Normal file
View File

@@ -0,0 +1,13 @@
# F.A.Q.
Nobody has *actually* asked these questions. But, we needed somewhere to put all the important bits and bobs that didn't fit anywhere else. So, here you go!
## What happens when the network goes down?
When a socket closed unexpectedly, all the commands that were already sent will reject as they might have been executed on the server. The rest will remain queued in memory until a new socket is established. If the client is closed—either by returning an error from [`reconnectStrategy`](./client-configuration.md#reconnect-strategy) or by manually calling `.disconnect()`—they will be rejected.
## How are commands batched?
Commands are pipelined using [`queueMicrotask`](https://nodejs.org/api/globals.html#globals_queuemicrotask_callback). Commands from the same "tick" will be sent in batches and respect the [`writableHighWaterMark`](https://nodejs.org/api/stream.html#stream_new_stream_writable_options).
If `socket.write()` returns `false`—meaning that ["all or part of the data was queued in user memory"](https://nodejs.org/api/net.html#net_socket_write_data_encoding_callback:~:text=all%20or%20part%20of%20the%20data%20was%20queued%20in%20user%20memory)—the commands will stack in memory until the [`drain`](https://nodejs.org/api/net.html#net_event_drain) event is fired.

30
docs/client-configuration.md Normal file
View File

@@ -0,0 +1,30 @@
# `createClient` configuration
| Property | Default | Description |
|--------------------------|------------------------------------------|------------------------------------------------------------------------------------------------------------------------------|
| socket | | Object defining socket connection properties |
| socket.url | | `[redis[s]:]//[[username][:password]@][host][:port]` |
| socket.host | `'localhost'` | Hostname to connect to |
| socket.port | `6379` | Port to connect to |
| socket.username | | ACL username ([see ACL guide](https://redis.io/topics/acl)) |
| socket.password | | ACL password or the old "--requirepass" password |
| socket.connectTimeout | `5000` | The timeout for connecting to the Redis Server (in milliseconds) |
| socket.noDelay | `true` | Enable/disable the use of [`Nagle's algorithm`](https://nodejs.org/api/net.html#net_socket_setnodelay_nodelay) |
| socket.keepAlive | `5000` | Enable/disable the [`keep-alive`](https://nodejs.org/api/net.html#net_socket_setkeepalive_enable_initialdelay) functionality |
| socket.tls | | Set to `true` to enable [TLS Configuration](https://nodejs.org/api/tls.html#tls_tls_connect_options_callback) |
| socket.reconnectStrategy | `retries => Math.min(retries * 50, 500)` | A function containing the [Reconnect Strategy](#reconnect-strategy) logic |
| modules | | Object defining which [Redis Modules](https://redis.io/modules) to include (TODO - document) |
| scripts | | Object defining Lua scripts to use with this client. See [Lua Scripts](../README.md#lua-scripts) |
| commandsQueueMaxLength | | Maximum length of the client's internal command queue |
| readonly | `false` | Connect in [`READONLY`](https://redis.io/commands/readonly) mode |
| legacyMode | `false` | Maintain some backwards compatibility (see the [Migration Guide](v3-to-v4.md)) |
| isolationPoolOptions | | See the [Isolated Execution Guide](./isolated-execution.md) |
## Reconnect Strategy
You can implement a custom reconnect strategy as a function that should:
- Receives the number of retries attempted so far.
- Should return `number | Error`:
- `number`: the time in milliseconds to wait before trying to reconnect again.
- `Error`: close the client and flush the commands queue.

67
docs/isolated-execution.md Normal file
View File

@@ -0,0 +1,67 @@
# Isolated Execution
Sometimes you want to run your commands on an exclusive connection. There are a few reasons to do this:
- You're using [transactions]() and need to `WATCH` a key or keys for changes.
- You want to run a blocking command that will take over the connection, such as `BLPOP` or `BLMOVE`.
- You're using the `MONITOR` command which also takes over a connection.
Below are several examples of how to use isolated execution.
> NOTE: Behind the scences we're using [`generic-pool`](https://www.npmjs.com/package/generic-pool) to provide a pool of connections that can be isolated. Go there to learn more.
## The Simple Secnario
This just isolates execution on a single connection. Do what you want with that connection:
```typescript
await client.executeIsolated(async isolatedClient => {
await isolatedClient.set('key', 'value');
await isolatedClient.get('key');
});
```
## Transactions
Things get a little more complex with transactions. Here we are `.watch()`ing some keys. If the keys change during the transaction, a `WatchError` is thrown when `.exec()` is called:
```typescript
try {
await client.executeIsolated(async isolatedClient => {
await isolatedClient.watch('key');
const multi = isolatedClient.multi()
.ping()
.get('key');
if (Math.random() > 0.5) {
await isolatedClient.watch('another-key');
multi.set('another-key', await isolatedClient.get('another-key') / 2);
}
return multi.exec();
});
} catch (err) {
if (err instanceof WatchError) {
// the transaction aborted
}
}
```
## Blocking Commands
For blocking commands, you can execute a tidy little one-liner:
```typescript
await client.executeIsolated(isolatedClient => isolatedClient.blPop('key'));
```
Or, you can just run the command directly, and provide the `isolated` option:
```typescript
await client.blPop(
commandOptions({ isolated: true }),
'key'
);
```

35
docs/v3-to-v4.md Normal file
View File

@@ -0,0 +1,35 @@
# v3 to v4 Migration Guide
Version 4 of Node Redis is a major refactor. While we have tried to maintain backwards compatibility where possible, several interfaces have changed. Read this guide to understand the differences and how to implement version 4 in your application.
## Breaking Changes
See the [Change Log](../CHANGELOG.md).
## Promises
Node Redis now uses native [Promises](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Promise) by default for all functions.
## Legacy Mode
Use legacy mode to preserve the backwards compatibility of commands while still getting access to the updated experience:
```typescript
const client = createClient({
legacyMode: true
});
// legacy mode
client.set('key', 'value', 'NX', (err, reply) => {
// ...
});
// version 4 interface is still accessible
await client.v4.set('key', 'value', {
NX: true
});
```
## `createClient`
The configuration object passed to `createClient` has changed significantly with this release. See the [client configuration guide](./client-configuration.md) for details.