CHECKPOINT_WARNING GUC variable addition

Started by Bruce Momjianover 23 years ago5 messagespatches
Jump to latest
#1Bruce Momjian
bruce@momjian.us

The following patch adds a GUC variable CHECKPOINT_WARNING which
generates a server log message if the checkpoint caused by WAL file
filling happens more frequently than (default) 30 seconds.

-- 
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 359-1001
  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073

Attachments:

/pgpatches/checkpointtext/plainDownload+47-0
#2Neil Conway
neilc@samurai.com
In reply to: Bruce Momjian (#1)
Re: CHECKPOINT_WARNING GUC variable addition

Bruce Momjian <pgman@candle.pha.pa.us> writes:

The following patch adds a GUC variable CHECKPOINT_WARNING which
generates a server log message if the checkpoint caused by WAL file
filling happens more frequently than (default) 30 seconds.

+    Checkpoints are fairly expensive because they force all dirty kernel
+    buffers to disk using the operating system <literal>sync()</> call.
+    Busy servers may fill checkpoint segment files too quickly,
+    causing excessive checkpointing. If such forced checkpoints happen
+    more than <varname>CHECKPOINT_WARNING</varname> seconds

That should be: "happen more [frequently] than ..."

Cheers,

Neil

--
Neil Conway <neilc@samurai.com> || PGP Key ID: DB3C29FC

#3Bruce Momjian
bruce@momjian.us
In reply to: Neil Conway (#2)
Re: CHECKPOINT_WARNING GUC variable addition

Neil Conway wrote:

Bruce Momjian <pgman@candle.pha.pa.us> writes:

The following patch adds a GUC variable CHECKPOINT_WARNING which
generates a server log message if the checkpoint caused by WAL file
filling happens more frequently than (default) 30 seconds.

+    Checkpoints are fairly expensive because they force all dirty kernel
+    buffers to disk using the operating system <literal>sync()</> call.
+    Busy servers may fill checkpoint segment files too quickly,
+    causing excessive checkpointing. If such forced checkpoints happen
+    more than <varname>CHECKPOINT_WARNING</varname> seconds

That should be: "happen more [frequently] than ..."

Great. New text:

+    Checkpoints are fairly expensive because they force all dirty kernel
+    buffers to disk using the operating system <literal>sync()</> call.
+    Busy servers may fill checkpoint segment files too quickly,
+    causing excessive checkpointing. If such forced checkpoints happen
+    more frequently than <varname>CHECKPOINT_WARNING</varname> seconds,
+    a message, will be output to the server logs recommending increasing
+    <varname>CHECKPOINT_SEGMENTS</varname>.
-- 
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 359-1001
  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073
#4Tom Lane
tgl@sss.pgh.pa.us
In reply to: Bruce Momjian (#1)
Re: CHECKPOINT_WARNING GUC variable addition

Bruce Momjian <pgman@candle.pha.pa.us> writes:

+    causing excessive checkpointing. If such forced checkpoints happen
+    more than <varname>CHECKPOINT_WARNING</varname> seconds, a message

You mean "more often than"?

regards, tom lane

#5Bruce Momjian
bruce@momjian.us
In reply to: Tom Lane (#4)
Re: CHECKPOINT_WARNING GUC variable addition

Tom Lane wrote:

Bruce Momjian <pgman@candle.pha.pa.us> writes:

+    causing excessive checkpointing. If such forced checkpoints happen
+    more than <varname>CHECKPOINT_WARNING</varname> seconds, a message

You mean "more often than"?

Yes, new text suggested by Neil Conway is:

+ more frequently than <varname>CHECKPOINT_WARNING</varname> seconds,

-- 
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 359-1001
  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073