database/postgres
1
0
Fork 0
mirror of https://github.com/postgres/postgres.git synced 2025-07-24 14:22:24 +03:00
Code Activity

Make pg_stat_io count IOs as bytes instead of blocks for some operations

Browse Source 
Currently in pg_stat_io view, IOs are counted as blocks of size
BLCKSZ.  There are two limitations with this design:
* The actual number of I/O requests sent to the kernel is lower because
I/O requests may be merged before being sent.  Additionally, it gives
the impression that all I/Os are done in block size, which shadows the
benefits of merging I/O requests.
* Some patches are under work to extend pg_stat_io for the tracking of
operations that may not be linked to the block size.  For example, WAL
read IOs are done in variable bytes and it is not possible to correctly
show these IOs in pg_stat_io view, and we want to keep all this data in
a single system view rather than spread it across multiple relations to
ease monitoring.

WaitReadBuffers() can now be tracked as a single read operation
worth N blocks.  Same for ExtendBufferedRelShared() and
ExtendBufferedRelLocal() for extensions.

Three columns are added to pg_stat_io for reads, writes and extensions
for the byte calculations.  op_bytes, which was always hardcoded to
BLCKSZ, is removed.  IO backend statistics are updated to reflect these
changes.

Bump catalog version.

Author: Nazir Bilal Yavuz
Reviewed-by: Bertrand Drouvot, Melanie Plageman
Discussion: https://postgr.es/m/CAN55FZ0oqxBaaHAEsj=xFqkzE3n5P=3RA1V_igXwL-RV7QRzyw@mail.gmail.com
This commit is contained in:
Michael Paquier
2025-01-14 12:14:29 +09:00
parent b4a07f532b
commit f92c854cf4
12 changed files with 170 additions and 77 deletions
Download Patch File Download Diff File Expand all files Collapse all files

86
src/backend/utils/adt/pgstatfuncs.c
View File

@ -1285,14 +1285,16 @@ typedef enum io_stat_col
IO_COL_OBJECT,
IO_COL_CONTEXT,
IO_COL_READS,
IO_COL_READ_BYTES,
IO_COL_READ_TIME,
IO_COL_WRITES,
IO_COL_WRITE_BYTES,
IO_COL_WRITE_TIME,
IO_COL_WRITEBACKS,
IO_COL_WRITEBACK_TIME,
IO_COL_EXTENDS,
IO_COL_EXTEND_BYTES,
IO_COL_EXTEND_TIME,
IO_COL_CONVERSION,
IO_COL_HITS,
IO_COL_EVICTIONS,
IO_COL_REUSES,
@ -1333,11 +1335,36 @@ pgstat_get_io_op_index(IOOp io_op)
pg_unreachable();
}
/*
* Get the number of the column containing IO bytes for the specified IOOp.
* If an IOOp is not tracked in bytes, IO_COL_INVALID is returned.
*/
static io_stat_col
pgstat_get_io_byte_index(IOOp io_op)
{
switch (io_op)
{
case IOOP_EXTEND:
return IO_COL_EXTEND_BYTES;
case IOOP_READ:
return IO_COL_READ_BYTES;
case IOOP_WRITE:
return IO_COL_WRITE_BYTES;
case IOOP_EVICT:
case IOOP_FSYNC:
case IOOP_HIT:
case IOOP_REUSE:
case IOOP_WRITEBACK:
return IO_COL_INVALID;
}
elog(ERROR, "unrecognized IOOp value: %d", io_op);
pg_unreachable();
}
/*
* Get the number of the column containing IO times for the specified IOOp.
* This function encodes our assumption that IO time for an IOOp is displayed
* in the view in the column directly after the IOOp counts. If an op has no
* associated time, IO_COL_INVALID is returned.
* If an op has no associated time, IO_COL_INVALID is returned.
*/
static io_stat_col
pgstat_get_io_time_index(IOOp io_op)
@ -1345,11 +1372,15 @@ pgstat_get_io_time_index(IOOp io_op)
switch (io_op)
{
case IOOP_READ:
return IO_COL_READ_TIME;
case IOOP_WRITE:
return IO_COL_WRITE_TIME;
case IOOP_WRITEBACK:
return IO_COL_WRITEBACK_TIME;
case IOOP_EXTEND:
return IO_COL_EXTEND_TIME;
case IOOP_FSYNC:
return pgstat_get_io_op_index(io_op) + 1;
return IO_COL_FSYNC_TIME;
case IOOP_EVICT:
case IOOP_HIT:
case IOOP_REUSE:
@ -1408,18 +1439,11 @@ pg_stat_io_build_tuples(ReturnSetInfo *rsinfo,
else
nulls[IO_COL_RESET_TIME] = true;
/*
* Hard-code this to the value of BLCKSZ for now. Future values
* could include XLOG_BLCKSZ, once WAL IO is tracked, and constant
* multipliers, once non-block-oriented IO (e.g. temporary file
* IO) is tracked.
*/
values[IO_COL_CONVERSION] = Int64GetDatum(BLCKSZ);
for (int io_op = 0; io_op < IOOP_NUM_TYPES; io_op++)
{
int op_idx = pgstat_get_io_op_index(io_op);
int time_idx = pgstat_get_io_time_index(io_op);
int byte_idx = pgstat_get_io_byte_index(io_op);
/*
* Some combinations of BackendType and IOOp, of IOContext and
@ -1436,19 +1460,39 @@ pg_stat_io_build_tuples(ReturnSetInfo *rsinfo,
else
nulls[op_idx] = true;
/* not every operation is timed */
if (time_idx == IO_COL_INVALID)
continue;
if (!nulls[op_idx])
{
PgStat_Counter time =
bktype_stats->times[io_obj][io_context][io_op];
/* not every operation is timed */
if (time_idx != IO_COL_INVALID)
{
PgStat_Counter time =
bktype_stats->times[io_obj][io_context][io_op];
values[time_idx] = Float8GetDatum(pg_stat_us_to_ms(time));
values[time_idx] = Float8GetDatum(pg_stat_us_to_ms(time));
}
/* not every IO is tracked in bytes */
if (byte_idx != IO_COL_INVALID)
{
char buf[256];
PgStat_Counter byte =
bktype_stats->bytes[io_obj][io_context][io_op];
/* Convert to numeric */
snprintf(buf, sizeof buf, UINT64_FORMAT, byte);
values[byte_idx] = DirectFunctionCall3(numeric_in,
CStringGetDatum(buf),
ObjectIdGetDatum(0),
Int32GetDatum(-1));
}
}
else
nulls[time_idx] = true;
{
if (time_idx != IO_COL_INVALID)
nulls[time_idx] = true;
if (byte_idx != IO_COL_INVALID)
nulls[byte_idx] = true;
}
}
tuplestore_putvalues(rsinfo->setResult, rsinfo->setDesc,