Re: Postgres Replaying WAL slowly - Mailing list pgsql-performance

From Jeff Frost
Subject Re: Postgres Replaying WAL slowly
Date
Msg-id 571D044B-260A-4E06-ACBC-8869B8F73C0F@pgexperts.com
Whole thread Raw
In response to Re: Postgres Replaying WAL slowly  (Jeff Frost <jeff@pgexperts.com>)
Responses Re: Postgres Replaying WAL slowly  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-performance

On Jun 30, 2014, at 1:46 PM, Jeff Frost <jeff@pgexperts.com> wrote:

So it seems like we have a candidate explanation.  I'm a bit surprised
that StandbyReleaseLocks would get this slow if there are only a dozen
AccessExclusiveLocks in place at any one time, though.  Perhaps that
was a low point and there are often many more?



Entirely possible that it was a low point.  We'll set up some monitoring to track the number of AccessExclusiveLocks and see how much variance there is throughout the day.


Since we turned on the monitoring for that, we had a peak of 13,550 AccessExclusiveLocks.  So far most of the samples have been in the double digit, with that and two other outliers: 6,118 and 12,747.

pgsql-performance by date:

Previous
From: Jeff Frost
Date:
Subject: Re: Postgres Replaying WAL slowly
Next
From: Tom Lane
Date:
Subject: Re: Postgres Replaying WAL slowly