From: Tom Lane <tgl@sss.pgh.pa.us>
To: Justin Pryzby <pryzby@telsasoft.com>
cc: padam.tehlan@leptonsoftware.com, pgsql-bugs@lists.postgresql.org
Subject: Re: BUG #17384: ERROR: missing chunk number 0 for toast value
 152073604 in pg_toast_2619
In-reply-to: <20220203044257.GH23027@telsasoft.com>
References: <17384-f50f2eedf541e512@postgresql.org>
 <20220203044257.GH23027@telsasoft.com>
Comments: In-reply-to Justin Pryzby <pryzby@telsasoft.com>
	message dated "Wed, 02 Feb 2022 22:42:57 -0600"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <198232.1643863841.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Wed, 02 Feb 2022 23:50:41 -0500
Message-ID: <198233.1643863841@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/198233.1643863841%40sss.pgh.pa.us>
Precedence: bulk

Justin Pryzby <pryzby@telsasoft.com> writes:
> On Thu, Jan 27, 2022 at 08:57:28PM +0000, PG Bug reporting form wrote:
>> Issue: ERROR: missing chunk number 0 for toast value 152073604 in
>> pg_toast_2619
>> we are facing issue with mentioned short description error while trying=
 any
>> operation on database. =


> From my notes from a handful of years ago, I think you'll have luck doin=
g some
> things like this:

The good news is that that's the toast table for pg_statistic,
which means that there is exactly no irreplaceable data in it.
If all else fails, you can just truncate pg_statistic and then
re-ANALYZE the whole database to rebuild it.

(I say "just" truncate pg_statistic, but you'll have to do that
in a standalone backend started with the -O option, else there
are permissions checks that will prevent it.)

The bad news is that there might be more corruption lurking
elsewhere.  What event precipitated this?

I think we have seen some bugs that would preferentially affect
pg_statistic, but not recently.  If you're not running a current
minor release, an update is advisable.

			regards, tom lane