Re: Quick-and-dirty compression for WAL backup blocks

From: Rod Taylor <pg(at)rbt(dot)ca>
To: "Jim C(dot) Nasby" <decibel(at)decibel(dot)org>
Cc: pgsql-hackers(at)postgresql(dot)org, Mark Cave-Ayland <m(dot)cave-ayland(at)webbased(dot)co(dot)uk>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Subject: Re: Quick-and-dirty compression for WAL backup blocks
Date: 2005-06-09 17:37:39
Message-ID: 1118338659.725.69.camel@home
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

> > I'd say that's an improvement worth having, especially considering that
> > it requires no net expenditure of CPU time. But the table is certainly
> > still open to discuss more complicated approaches.
>
> If it's not hard to hack in as a test, it'd be interesting to see what
> additional gains a more aggresive compression algorithm like LZW got.
> CPU is more of a concern in that case, but for databases generating a
> lot of WAL it might still be a win.

I've generate a fair amount of WAL segments (about 20GB per day), and
have a CPU issue.

I implemented a cronjob which compresses them using gzip on a different
machine.

Any chance we could have an official compression tool which is
independent of the server itself so we can distribute the load a little?

> BTW, is this the thread you reffered to? I wasn't able to find it in the
> TODO and had to go into the archives to find it...
> http://archives.postgresql.org/pgsql-performance/2005-04/msg00264.php
--

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Bruce Momjian 2005-06-09 18:00:03 Re: [HACKERS] Added schema selection to pg_restore
Previous Message Marc G. Fournier 2005-06-09 17:28:46 Re: The Contrib Roundup (long)