Tom Lane wrote:
> Adam Witney <awitney@sgul.ac.uk> writes:
>> I don't know, so posting your question to the list. I meant to ask last
>> night in fact, should i be worried as to why this occurred?
>
> Yeah, you should, but since you hadn't given us any context about the
> problem I figured you only cared about a quick fix. Otherwise I would
> have suggested saving a physical copy of the broken index and table
> for analysis.
>
> The symptoms imply that some of the index entries were pointing to rows
> that didn't belong to them (ie, rows with a different bioassay_id value
> than what the index entry said). The only ways I know of to explain
> that involve database or system crashes ... you have any recently?
Hi, yes, i was a bit worried about it, and hence went for the fix first!
I don't think there have been any crashes, the uptime is showing 189
days and i don't see anything obvious in the PostgreSQL log. The data
rows that were incorrectly being pointed to were only uploaded on friday
and there had been zero user connections since friday evening. i do run
a Vacuum Analyse and a full dump every night, i don't know if that makes
a difference though. I have had problems with this table before (you
probably don't remember) that required me (with your help Tom) to zero
out blocks and reload some data.
Anyway i will try to get a copy of the index if it happens again
thanks again for your help
adam
--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.