From: | Matthew Peter <survivedsushi(at)yahoo(dot)com> |
---|---|
To: | Richard Huxton <dev(at)archonet(dot)com> |
Cc: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: transaction toggling |
Date: | 2005-10-05 17:00:15 |
Message-ID: | 20051005170015.99969.qmail@web35211.mail.mud.yahoo.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
--- Richard Huxton <dev(at)archonet(dot)com> wrote:
> Matthew Peter wrote:
> > Is there a way to disable transactions on certian
> > queries? I read a while back that this wasn't
> > possible, just wondering if it is in 8.1 or
> planned in
> > the future?
> >
> > Does it even make a dent in the performance if a
> query
> > is wrapped in a transaction instead of out? Such
> as a
> > view counter, like...
> >
> > update views = views + 1 where x = 1;?
>
> It's not in general possible to have part of your
> database under
> transactional control and part not.
>
> For example, what happens in your update if there
> are two processes
> incrementing the view counter - A is in a
> transaction, B is not.
> A begins transaction
> A increments view count
> B increments view count
> A rolls back transaction
>
> What should happen here?
I'm not worried about that. I know what transactions
are suppose to 'prevent', but for a hit counter I
don't need transaction support to rollback.
Having a busy site I don't see the benefits to
automatically wrap every transaction with BEGIN; ...
COMMIT; w/o option to IGNORE; ... DONE;
I don't want to consume more resources than necessary,
generating unnecessary overhead and junk for VACUUM to
clean up if possible...
MP
__________________________________
Yahoo! Mail - PC Magazine Editors' Choice 2005
http://mail.yahoo.com
From | Date | Subject | |
---|---|---|---|
Next Message | Berend Tober | 2005-10-05 17:11:45 | Re: Securing Postgres |
Previous Message | Richard_D_Levine | 2005-10-05 16:53:36 | PostgreSQL on Slash Dot |