Re: Multixacts wraparound monitoring - Mailing list pgsql-general

From Pavlov, Vladimir
Subject Re: Multixacts wraparound monitoring
Date
Msg-id B38B34EC5621E34DABCE13E8B18936E602DC2C14BE58@EXSERV.Gallup.tns
Whole thread Raw
In response to Re: Multixacts wraparound monitoring  (Adrian Klaver <adrian.klaver@aklaver.com>)
Responses Re: Multixacts wraparound monitoring
List pgsql-general
Thanks for your reply.
Yes, the first thing I looked at the statistics from pg_stat_activity.
But I have a transaction is not more than 60 seconds and the condition 'idle in transaction' lasts only a few seconds.

Kind regards,
 
Vladimir Pavlov

-----Original Message-----
From: Adrian Klaver [mailto:adrian.klaver@aklaver.com]
Sent: Thursday, March 24, 2016 4:36 PM
To: Pavlov Vladimir; pgsql-general@postgresql.org
Subject: Re: [GENERAL] Multixacts wraparound monitoring

On 03/24/2016 12:54 AM, Pavlov, Vladimir wrote:
> Hello,
>
> How can we determine when an error of approximation multixacts wraparound?
>
> According to the information from pg_class:
>
> select datname,datminmxid from pg_database;
>
>      datname     | datminmxid
>
> ----------------+------------
>
> template1      |  347462426
>
> template0      |  347462426
>
> postgres          |  347462426
>
> zabbix             |  467261307
>
> db_3                |  291141939
>
> db_1               |  388282963
>
> db                    |  388282963
>
> But when the vacuum/autovacuum starts up, an error occurs:
>
> WARNING:  oldest multixact is far in the past
>
> HINT:  Close open transactions with multixacts soon to avoid
> wraparound problems.

The above would seem to be the key. Take a look at what is in:

select * from pg_stat_activity;

You are looking for long running queries and/or 'idle in transaction'
queries'.

For more information see:

http://www.postgresql.org/docs/9.3/static/monitoring-stats.html#PG-STAT-ACTIVITY-VIEW

>
> If I understand correctly, approaching Multixact member wraparound.
>
> But how to understand when it comes exactly and what to do?
>
> PostgreSQL version - 9.3.10, OS Debian 7.8.
>
> Thank you.
>
> Sorry, if I chose the wrong mailing list.
>
> Kind regards,
>
> *Vladimir Pavlov*
>


--
Adrian Klaver
adrian.klaver@aklaver.com


pgsql-general by date:

Previous
From: "David G. Johnston"
Date:
Subject: Re: Only owners can ANALYZE tables...seems overly restrictive
Next
From: Howard News
Date:
Subject: Uninstalled working db by mistake