Re: Fragged State in 7.0.2 - Mailing list pgsql-hackers

From Tim Perdue
Subject Re: Fragged State in 7.0.2
Date
Msg-id 39B65F62.D8E051F5@valinux.com
Whole thread Raw
In response to Re: Fragged State in 7.0.2  ("Mike Mascari" <mascarm@mascari.com>)
List pgsql-hackers
It doesn't suprise me that it doesn't work, but I am surprised to get
into a half-baked state if an abort does happen for some reason.

NOTICE:  Caution: DROP TABLE cannot be rolled back, so don't abort now

Probably what should happen is it should error out if you try to do
things like this in a transaction, or autocommit for you no matter what.
Whatever it takes to not be in a half-baked state.

Tim


Tom Lane wrote:
> 
> "Mike Mascari" <mascarm@mascari.com> writes:
> > Rolling back DDL statements properly
> > in a MVCC transaction environment is very difficult, as
> > you can imagine. IIRC Oracle cheats, Informix and DEC Rdb
> > lock the DDL target until transaction commit, etc. If
> > the PostgreSQL team implements their stated goal in this
> > area, it will be far superior to its commercial counterparts.
> 
> AFAIK we intend to keep the current behavior of exclusively
> locking any table you try to drop or modify.  So it'll be
> pretty much like the Informix/RDB behavior.
> 
> But yes, at the moment DROP or RENAME inside a transaction is
> pretty risky (and 7.0 tells you so, with an annoying NOTICE).
> 
>                         regards, tom lane

-- 
Founder - PHPBuilder.com / Geocrawler.com
Lead Developer - SourceForge
VA Linux Systems
408-542-5723


pgsql-hackers by date:

Previous
From: Tom Samplonius
Date:
Subject: Row versioning in the ODBC driver...
Next
From: Alexey Raschepkin
Date:
Subject: 'select pg_class from pg_class' error