Re: BUG #16593: pg_upgrade make corrupt UK indexes - Mailing list pgsql-bugs

From Bruce Momjian
Subject Re: BUG #16593: pg_upgrade make corrupt UK indexes
Date
Msg-id 20200826225730.GD13370@momjian.us
Whole thread Raw
In response to Re: BUG #16593: pg_upgrade make corrupt UK indexes  (Magnus Hagander <magnus@hagander.net>)
Responses Re: BUG #16593: pg_upgrade make corrupt UK indexes  (László Tóth <hali19790320@gmail.com>)
List pgsql-bugs
On Wed, Aug 26, 2020 at 09:49:00AM +0200, Magnus Hagander wrote:
> 
> 
> On Wed, Aug 26, 2020 at 7:38 AM Nick Cleaton <nick@cleaton.net> wrote:
> 
>     On Tue, 25 Aug 2020 at 23:55, PG Bug reporting form <noreply@postgresql.org
>     > wrote:
> 
>         We upgrade from 9.6 Centos7 to 12.3 Centos 8
> 
> 
>     This from https://www.postgresql.org/docs/12/amcheck.html#id-1.11.7.11.9
>     might be relevent:
> 
>     > Comparisons of datums of a collatable type like text must be immutable
>     (just as all comparisons used for B-Tree index scans must be immutable),
>     which implies that operating system collation rules must never change.
>     Though rare, updates to operating system collation rules can cause these
>     issues. More commonly, an inconsistency in the collation order between a
>     master server and a standby server is implicated, possibly because the 
>     major operating system version in use is inconsistent
> 
> 
> And in particular also see https://wiki.postgresql.org/wiki/Locale_data_changes
> , which shows that the step from Centos7 to Centos8 is one that causes this
> problem. You need to reindex all text/varchar indexes when you  make that
> upgrade.

And we should have a script ready for people to do this.  :-)

-- 
  Bruce Momjian  <bruce@momjian.us>        https://momjian.us
  EnterpriseDB                             https://enterprisedb.com

  The usefulness of a cup is in its emptiness, Bruce Lee




pgsql-bugs by date:

Previous
From: Alvaro Herrera
Date:
Subject: Re: ALTER TABLE ALTER COLUMN SET TYPE crash
Next
From: Michael Paquier
Date:
Subject: Re: BUG #16594: DROP INDEX CONCURRENTLY fails on partitioned table with a non helpful error message.