Re: found xmin from before relfrozenxid on pg_catalog.pg_authid - Mailing list pgsql-general

From Andres Freund
Subject Re: found xmin from before relfrozenxid on pg_catalog.pg_authid
Date
Msg-id 20180522193007.4bi5oluqb7c72oq2@alap3.anarazel.de
Whole thread Raw
In response to Re: found xmin from before relfrozenxid on pg_catalog.pg_authid  (Maxim Boguk <maxim.boguk@gmail.com>)
Responses Re: found xmin from before relfrozenxid on pg_catalog.pg_authid  (Maxim Boguk <maxim.boguk@gmail.com>)
Re: found xmin from before relfrozenxid on pg_catalog.pg_authid  (Maxim Boguk <maxim.boguk@gmail.com>)
List pgsql-general
Hi,

On 2018-05-22 22:18:15 +0300, Maxim Boguk wrote:
> On Tue, May 22, 2018 at 9:47 PM, Andres Freund <andres@anarazel.de> wrote:
> > >  select relfrozenxid from pg_class where relname='pg_authid';
> > >  relfrozenxid
> > > --------------
> > >    2863429136

> ​select txid_current();
>  txid_current
> --------------
>   41995913769

So that's an xid of 3341208114, if you leave the epoch out. What's
autovacuum_freeze_max_age set to in that cluster?

Can you show pg_controldata output, and relminmxid from that cluster?

I might be daft here, but it's surely curious that the relfrozenxid from
the error and pg_catalog are really different (number of digits):
catalog: 2863429136
error:    248712603


> ​About gdb bt - it's tricky because it is mission critical master db of
> huge project.
> I'll will try promote backup replica and check is issue persist there and
> if yes - we will have our playground for a while, but it will require
> sometime to arrange.​

You should be ok to just bt that in the running cluster, but I
definitely understand if you don't want to do that...  I'd appreciate if
you set up the a playground, because this seems like something that'll
reappear.

Greetings,

Andres Freund


pgsql-general by date:

Previous
From: Tiffany Thang
Date:
Subject: pg_multixact/members growing
Next
From: Maxim Boguk
Date:
Subject: Re: found xmin from before relfrozenxid on pg_catalog.pg_authid