| From: | Gregory Stark <stark(at)enterprisedb(dot)com> |
|---|---|
| To: | "Alvaro Herrera" <alvherre(at)commandprompt(dot)com> |
| Cc: | "Shane Ambler" <pgsql(at)Sheeky(dot)Biz>, "PostgreSQL Hackers" <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: auto-vacuum conflicting with long running analyze |
| Date: | 2008-04-05 19:06:31 |
| Message-ID: | 874pagcea0.fsf@oxford.xeocode.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
"Alvaro Herrera" <alvherre(at)commandprompt(dot)com> writes:
> Shane Ambler wrote:
>
>> Given that the analyze will obviously take a long time, is this scenario
>> likely to happen with 8.3.1? or has it been fixed since 8.1.x?
>
> In 8.3, autovacuum cancels itself if it sees it is conflicting with
> another query.
But that's because it could *block* other queries. Not because it should be
causing deadlocks. Unless you're doing analyze within a transaction I don't
see why a autovacuum would be causing deadlocks. There are useful reasons to
run analyze within a transaction but I suspect you're not actually doing it
for any of those reasons.
--
Gregory Stark
EnterpriseDB http://www.enterprisedb.com
Ask me about EnterpriseDB's 24x7 Postgres support!
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Alvaro Herrera | 2008-04-05 20:49:19 | Re: [HACKERS] Patch queue -> wiki |
| Previous Message | Shane Ambler | 2008-04-05 18:37:15 | Re: auto-vacuum conflicting with long running analyze |