Re: hung backends stuck in spinlock heavy endless loop - Mailing list pgsql-hackers

From Peter Geoghegan
Subject Re: hung backends stuck in spinlock heavy endless loop
Date
Msg-id CAM3SWZRicfLszVTy0MnP+UGiM3MFXOOmqFkR6DcfUz29PfaU4g@mail.gmail.com
Whole thread Raw
In response to Re: hung backends stuck in spinlock heavy endless loop  (Peter Geoghegan <pg@heroku.com>)
List pgsql-hackers
On Wed, Jan 14, 2015 at 5:23 PM, Peter Geoghegan <pg@heroku.com> wrote:
> My immediate observation here is that blocks 2 and 9 have identical
> metadata (from their page opaque area), but partially non-matching
> data items (however, the number of items on each block is consistent
> and correct according to that metadata, as far as it goes). I think
> block 9 is supposed to have a right-link to block 5 (block 5 seems to
> think so, at least -- its left link is 9).

For the convenience of others, here is a version with a normalized
data column added (interpreting the datums as little-endian Oids).

--
Peter Geoghegan

Attachment

pgsql-hackers by date:

Previous
From: Alvaro Herrera
Date:
Subject: Re: Safe memory allocation functions
Next
From: Alvaro Herrera
Date:
Subject: Re: Improving RLS qual pushdown