Re: Using streaming replication as log archiving

From: Magnus Hagander <magnus(at)hagander(dot)net>
To: Kevin Grittner <Kevin(dot)Grittner(at)wicourts(dot)gov>
Cc: Aidan Van Dyk <aidan(at)highrise(dot)ca>, PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Using streaming replication as log archiving
Date: 2010-10-01 08:08:08
Message-ID: AANLkTik=Zy5XDsiEuJ756EAcNCH2Qev2ws9Mcb+gKPXQ@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Thu, Sep 30, 2010 at 17:25, Kevin Grittner
<Kevin(dot)Grittner(at)wicourts(dot)gov> wrote:
> Aidan Van Dyk <aidan(at)highrise(dot)ca> wrote:
>
>> When the "being written to" segmnt copmletes moves to the final
>> location, he'll get an extra whole "copy" of the file.  But of the
>> "move" can be an exec of his scritpt, the compressed/gzipped final
>> result shouldn't be that bad.  Certainly no worse then what he's
>> currently getting with archive command ;-)  And he's got the
>> uncompressed incimental updates as they are happening.
>
> Hmmm...  As long as streaming replication doesn't send the "tail" of
> an incomplete WAL segment file, the only thing we'd be missing on
> the send to the central location is the compression.  That's
> typically reducing the size of the transmission by 50% to 75% (e.g.,
> the gzipped "full" files are usually in the range of 4MB to 8MB).
> At our WAN speeds, that is significant.  I don't suppose that
> streaming replication uses (or offers as an option) a compressed
> stream?

No, it sends a regular COPY stream with the raw transaction log data.
So the files generated will eventually exactly 16Mb, just like they
are in pg_xlog.

--
 Magnus Hagander
 Me: http://www.hagander.net/
 Work: http://www.redpill-linpro.com/

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Hitoshi Harada 2010-10-01 08:17:19 Re: wip: functions median and percentile
Previous Message Magnus Hagander 2010-10-01 08:07:07 Re: Patch author name on commitfest page