Re: tuple concurrently updated

[Evan Bauer <evanbauer@xxxxxxx>]

Walter,

I agree with your diagnosis of a database corruption, the error message from your attempt to dump essentially provides it:

pg_dump: Fehlermeldung vom Server: FEHLER:  unexpected chunk number 0
(expected 1) for toast value 1261719035 in pg_toast_1340113

With all of the TEXT columns in the table, a corrupted TOAST pointer or (more likely) TOAST table page looks to be the culprit.

I agree that a rebuild seems to be the best (though lengthy) approach.

Your post here was my first introduction to Open Street Map — a fascinating project — thanks.

Cheers,

- Evan

Evan Bauer
eb@xxxxxxxxxxxxx
+1 646 641 2973
Skype: evanbauer

On Aug 28, 2018, at 09:58, Walter Nordmann <walter.nordmann@xxxxxxxxx> wrote:

updates are done using a program, that is without my control: osm2pgsql.
this program does everthing which is necessary including locking. and
it't running for years on many, many openstreetmap servers.

because there was a crash before, it must be a data corruption problem,
not programming error.

regards

walter



Am 28.08.2018 um 15:45 schrieb pavan95:

Hi Walter,

Why can't you go with a postgres service restart in order to get rid of this
error?

DELETE FROM planet_osm_line WHERE osm_id = -166570;
FEHLER:  tuple concurrently updated

Concurrent transactions can update or delete one or more of those rows
before DELETE can lock the rows (at least with the default isolation level
READ COMMITTED). This would result in your error message.

To defend against this race condition, lock the rows in the SELECT with FOR
UPDATE (or other options)

Regards,
Pavan



--
Sent from: http://www.postgresql-archive.org/PostgreSQL-admin-f2076596.html