pg.ddx.io pgsql-admin@postgresql.org mailing list archive
help / color / mirror / Atom feedRecovery conflict due to buffer pins
3+ messages / 2 participants
[nested] [flat]
* Recovery conflict due to buffer pins
@ 2022-06-13 18:05 Nikhil Shetty <nikhil.dba04@gmail.com>
0 siblings, 1 reply; 3+ messages in thread
From: Nikhil Shetty @ 2022-06-13 18:05 UTC (permalink / raw)
To: Pgsql-admin <pgsql-admin@lists.postgresql.org>
Hi,
PostgreSQL version - 11.7
We are seeing using few queries to fetch data from standby for monitoring
but we often see that the query is cancelled due to 'conflict
with recovery'.
Example:
ERROR: canceling statement due to *conflict with recovery*
2022-06-12 04:30:01 UTC [3594]: [5-1]
user=<user>,db=postgres,app=<app>,client=127.0.0.1DETAIL: User was holding
shared buffer pin for too long.
2022-06-12 04:30:01 UTC [3594]: [6-1]
user=<user>,db=postgres,app=<app>,client=127.0.0.1STATEMENT: SET
statement_timeout = 30000; SELECT 'dbsize_bytes',
COALESCE(sum(pg_database_size(datname)),0) FROM pg_stat_database where
datname NOT IN ('postgres','template0','template1');
2022-06-12 04:30:01 UTC [3324]: [3-1]
user=<user>,db=postgres,app=<app>,client=127.0.0.1ERROR: canceling
statement due to *conflict with recovery*
04:30:01 UTC [3324]: [4-1]
user=<user>,db=postgres,app=<app>,client=127.0.0.1DETAIL: User was holding
shared buffer pin for too long.
04:30:01 UTC [3324]: [5-1]
user=<user>,db=postgres,app=<app>,client=127.0.0.1STATEMENT: SELECT
'tuple_returned', sum(tup_returned) FROM pg_stat_database;
I have verified that there is a vacuum operation happening at the same time
on primary.
We have set below parameters in standby
hot_standby_feedback = on
max_standby_streaming_delay = 0
How can we avoid this error on standby?
Regards,
Nikhil
^ permalink raw reply [nested|flat] 3+ messages in thread
* Re: Recovery conflict due to buffer pins
@ 2022-06-13 18:30 Tom Lane <tgl@sss.pgh.pa.us>
parent: Nikhil Shetty <nikhil.dba04@gmail.com>
0 siblings, 1 reply; 3+ messages in thread
From: Tom Lane @ 2022-06-13 18:30 UTC (permalink / raw)
To: Nikhil Shetty <nikhil.dba04@gmail.com>; +Cc: Pgsql-admin <pgsql-admin@lists.postgresql.org>
Nikhil Shetty <nikhil.dba04@gmail.com> writes:
> 2022-06-12 04:30:01 UTC [3324]: [3-1]
> user=<user>,db=postgres,app=<app>,client=127.0.0.1ERROR: canceling
> statement due to *conflict with recovery*
> 04:30:01 UTC [3324]: [4-1]
> user=<user>,db=postgres,app=<app>,client=127.0.0.1DETAIL: User was holding
> shared buffer pin for too long.
> We have set below parameters in standby
> hot_standby_feedback = on
> max_standby_streaming_delay = 0
> How can we avoid this error on standby?
Use a larger max_standby_streaming_delay. Setting it to zero means
precisely that conflicting queries will be canceled immediately.
regards, tom lane
^ permalink raw reply [nested|flat] 3+ messages in thread
* Re: Recovery conflict due to buffer pins
@ 2022-06-14 11:58 Nikhil Shetty <nikhil.dba04@gmail.com>
parent: Tom Lane <tgl@sss.pgh.pa.us>
0 siblings, 0 replies; 3+ messages in thread
From: Nikhil Shetty @ 2022-06-14 11:58 UTC (permalink / raw)
To: Tom Lane <tgl@sss.pgh.pa.us>; +Cc: Pgsql-admin <pgsql-admin@lists.postgresql.org>
Hi Tom,
We have set max_standby_streaming_delay to 0 because we have a synchronous
standby with synchronous_commit as 'remote_apply'. We want to make sure
that queries are cancelled so that it doesn't affect the primary.
Is there any other way to resolve this apart from setting
max_standby_streaming_delay
to a value greater than zero?
Thanks,
Nikhil
On Tue, Jun 14, 2022 at 12:00 AM Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Nikhil Shetty <nikhil.dba04@gmail.com> writes:
> > 2022-06-12 04:30:01 UTC [3324]: [3-1]
> > user=<user>,db=postgres,app=<app>,client=127.0.0.1ERROR: canceling
> > statement due to *conflict with recovery*
> > 04:30:01 UTC [3324]: [4-1]
> > user=<user>,db=postgres,app=<app>,client=127.0.0.1DETAIL: User was
> holding
> > shared buffer pin for too long.
>
> > We have set below parameters in standby
> > hot_standby_feedback = on
> > max_standby_streaming_delay = 0
>
> > How can we avoid this error on standby?
>
> Use a larger max_standby_streaming_delay. Setting it to zero means
> precisely that conflicting queries will be canceled immediately.
>
> regards, tom lane
>
^ permalink raw reply [nested|flat] 3+ messages in thread
end of thread, other threads:[~2022-06-14 11:58 UTC | newest]
Thread overview: 3+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2022-06-13 18:05 Recovery conflict due to buffer pins Nikhil Shetty <nikhil.dba04@gmail.com>
2022-06-13 18:30 ` Tom Lane <tgl@sss.pgh.pa.us>
2022-06-14 11:58 ` Nikhil Shetty <nikhil.dba04@gmail.com>
This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox