Re: Reduction in WAL for UPDATEs - Mailing list pgsql-hackers

From Florian G. Pflug
Subject Re: Reduction in WAL for UPDATEs
Date
Msg-id 460A72F1.6080909@phlo.org
Whole thread Raw
In response to Re: Reduction in WAL for UPDATEs  (Kenneth Marshall <ktm@rice.edu>)
Responses Re: Reduction in WAL for UPDATEs  ("Simon Riggs" <simon@2ndquadrant.com>)
List pgsql-hackers
Kenneth Marshall wrote:
> We use DSPAM as one of our anti-spam options. Its UPDATE pattern is to
> increment a spam counter or a not-spam counter while keeping the user and
> token information the same. This would benefit from this optimization.
> Currently we are forced to use MySQL with MyISM tables to support the
> update load, although PostgreSQL 8.2 performance is right at the I/O
> break-even point for switching databases. With HOT and more optimized
> UPDATE I/O, 8.3 would give us enough I/O headroom to switch to PostgreSQL.

Interesting. I've switched from MySQL to PostgreSQL for dspam, because
of concurrency issues with MyISAM which caused bad performance.

I am eager to see how much HOT speeds of my setup, though ;-)

BTW, the "COMMIT NOWAIT" feature Simon Riggs proposed should provide
a huge speedup too, since dspam runs one transaction for each token
it has to update.

greetings, Florian Pflug


pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: Reduction in WAL for UPDATEs
Next
From: Tom Lane
Date:
Subject: Re: Reduction in WAL for UPDATEs