checkpoint_warning

A configuration parameter specifying the interval to warn about too-frequent checkpoints

checkpoint_warning is a configuration parameter determining the interval at which to write log messages if checkpoints which are due to WAL file growth occur too frequently.

checkpoint_warning was added in PostgreSQL 7.4.

Default value

The default value for checkpoint_warning is: 30s.

No warnings will be logged if checkpoint_timeout is less than checkpoint_warning, or if checkpoint_warning is set to 0.

Change history

Examples

Log excerpt from an operation (in this case a bulk data load) which is generating more checkpoints than usual:

[2021-03-29 10:49:29 UTC]    LOG:  00000: checkpoint starting: wal
...
[2021-03-29 10:49:34 UTC]    LOG:  00000: checkpoint complete: wrote 8480 buffers (51.8%); 0 WAL file(s) added, 0 removed, 33 recycled; write=5.037 s, sync=0.001 s, total=5.043 s; sync files=0, longest=0.000 s, average=0.000 s; distance=538195 kB, estimate=538195 kB
[2021-03-29 10:49:34 UTC]    LOG:  00000: checkpoints are occurring too frequently (5 seconds apart)
[2021-03-29 10:49:34 UTC]    HINT:  Consider increasing the configuration parameter "max_wal_size".
[2021-03-29 10:49:34 UTC]    LOG:  00000: checkpoint starting: wal
...
[2021-03-29 10:49:38 UTC]    LOG:  00000: checkpoint complete: wrote 807 buffers (4.9%); 0 WAL file(s) added, 3 removed, 30 recycled; write=3.906 s, sync=0.001 s, total=3.919 s; sync files=0, longest=0.000 s, average=0.000 s; distance=540689 kB, estimate=540689 kB
[2021-03-29 10:49:38 UTC]    LOG:  00000: checkpoints are occurring too frequently (4 seconds apart)
[2021-03-29 10:49:38 UTC]    HINT:  Consider increasing the configuration parameter "max_wal_size".
[2021-03-29 10:49:38 UTC]    LOG:  00000: checkpoint starting: wal
...
[2021-03-29 10:49:39 UTC]    LOG:  00000: checkpoint complete: wrote 384 buffers (2.3%); 0 WAL file(s) added, 3 removed, 30 recycled; write=0.394 s, sync=0.001 s, total=0.430 s; sync files=0, longest=0.000 s, average=0.000 s; distance=549825 kB, estimate=549825 kB
[2021-03-29 10:49:39 UTC]    LOG:  00000: checkpoints are occurring too frequently (1 second apart)
[2021-03-29 10:49:39 UTC]    HINT:  Consider increasing the configuration parameter "max_wal_size".

Categories

Checkpoints, GUC configuration item

See also

log_checkpoints, checkpoint_timeout, max_wal_size