Re: Errors while vacuuming large tables

On Mon, 14 Oct 2002 11:13:46 -0400, Jeff Boes wrote:

> ERROR:  cannot open segment 1 of relation table_D (target block
> 2337538109): No such file or directory
>
> and for table B:
>
> NOTICE:  Child itemid in update-chain marked as unused - can't continue
> repair_frag
> ERROR:  cannot open segment 3 of relation pg_toast_51207070 (target
> block 2336096317): No such file or directory
>

Well, we attempted a VACUUM FULL on these tables again, but got the
dreaded:

NOTICE:  Message from PostgreSQL backend:
        The Postmaster has informed me that some other backend
        died abnormally and possibly corrupted shared memory.
        I have rolled back the current transaction and am
        going to terminate your database system connection and exit.
        Please reconnect to the database system and repeat your query.

We also had two load spikes several hours after this occurred, where the
system was unusable for 5 minutes (this has also happened before).  We
suspect there was some corruption in the indexes.  I did a REINDEX on
each affected table followed by a VACUUM FULL, and this time the VACUUMs
finished without error.  All this may be superstition on my part, but I
report it here in case anyone recognizes the situation, or is as
desperate as I was to correct the problem ...

We are definitely moving to 7.2.3 ASAP, but it will be a week or two
before we can schedule this.


--
Jeff Boes                                      vox 616.226.9550 ext 24
Database Engineer                                     fax 616.349.9076
Nexcerpt, Inc.                                 http://www.nexcerpt.com
           ...Nexcerpt... Extend your Expertise