Subject: | (forw) [greg@turnstep.com: Re: DBD::Pg pg_cancel] |
Date: | Wed, 3 Mar 2010 11:25:11 -0500 |
To: | bug-DBD-Pg [...] rt.cpan.org |
From: | Stephen Frost <sfrost [...] snowman.net> |
Reporting bug that pg_cancel() unconditionally issues a rollback, even
tho the use might be addressing the issue with savepoints already.
Commentary from Greg below.
----- Forwarded message from Greg Sabino Mullane <greg@turnstep.com> -----
Date: Wed, 3 Mar 2010 16:07:28 -0000
From: Greg Sabino Mullane <greg@turnstep.com>
To: sfrost@snowman.net
X-Mailer: JoyMail 2.4.0
X-Spam-Status: No, score=-2.6 required=5.0 tests=AWL,BAYES_00 autolearn=ham
version=3.2.5
Subject: Re: DBD::Pg pg_cancel
-----BEGIN PGP SIGNED MESSAGE-----
Hash: RIPEMD160
Show quoted text
> Does DBD::Pg really need to issue a rollback when pg_cancel() is used?
> Maybe I missed something, but psql certainly doesn't appear to do
> that, and when you're using savepoints, I don't really see why
> pg_cancel should be issueing a rollback.
>
> Can we have DBD::Pg either changed to never issue a rollback, or get
> it configurable?
*blink* *checks code*. Yeah, from a quick glance, I think you are right.
The rollback is a convienence, but doesn't account for ROLLBACK TO.
Probably should get added as an official DBD::Pg bug, so it doesn't
fall through the cracks.
At this point, it's either that way for a good reason that we haven't
come across yet, or it's a mistake and can be taken out entirely, so
I don't think it needs to be configurable.
- --
Greg Sabino Mullane greg@turnstep.com
End Point Corporation http://www.endpoint.com/
PGP Key: 0x14964AC8 201003031106
http://biglumber.com/x/web?pk=2529DF6AB8F79407E94445B4BC9B906714964AC8
-----BEGIN PGP SIGNATURE-----
iEYEAREDAAYFAkuOiSYACgkQvJuQZxSWSsh8yQCg6wpiCd+lyMCuftIkYYzUraK/
1xEAn0WpAgznUsNkigxACs32JvlTHhWj
=F8PZ
-----END PGP SIGNATURE-----
Show quoted text----- End forwarded message -----
Message body not shown because it is not plain text.