Re: WAL Crash during index vacuuming 7.1beta4 - Mailing list pgsql-hackers

From Vadim Mikheev
Subject Re: WAL Crash during index vacuuming 7.1beta4
Date
Msg-id 004601c09009$6512a500$4e79583f@sectorbase.com
Whole thread Raw
In response to WAL Crash during index vacuuming 7.1beta4  (Giuseppe Tanzilli - CSF <g.tanzilli@gruppocsf.com>)
List pgsql-hackers
> during the nightly vacuum pgsql closed and do not start any more.
> Attached the log.
> 
> Seems the problem was rebuilding an Index,
> There is a way to force wal to ignore indexes ?

The problem was in redoing tuple movement in *table*.

> Can I delete it ?
>
...
>
> DEBUG:  redo starts at (6, 4144414796)
> NOTICE:  PageAddItem: tried overwrite of used ItemId
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> FATAL 2:  heap_update_redo: failed to add tuple

I think that I've just fixed this problem (must not check itemid'
flag in PageAddItem in overwrite mode when offset number
== maxoff + 1). I hope that Giuseppe will check new code soon.

Thanks to Giuseppe for help!

Vadim




pgsql-hackers by date:

Previous
From: Ryan Kirkpatrick
Date:
Subject: Re: [PATCHES] A Sparc/Linux patch (for 7.1), and a Linux rc.d/init.d script....
Next
From: "sourabh dixit"
Date:
Subject: Postgre SQL for Windows