Re: New statistics for WAL buffer dirty writes

2012/07/07 22:07, Euler Taveira wrote:
> On 07-07-2012 09:00, Satoshi Nagayasu wrote:
>> I've created new patch to get/reset statistics of WAL buffer
>> writes (flushes) caused by WAL buffer full.
>>
> This new statistic doesn't solve your problem (tune wal_buffers). It doesn't
> give you the wal_buffers value. It only says "hey, I needed more buffers so I
> write those dirty ones". It doesn't say how many. I would like to have
> something that says "hey, you have 1000 buffers available and you are using
> 100 buffers (10%)". This new statistic is only useful for decreasing the
> WALWriteLock contention.

I agree with that it would not tell the exact number for wal_buffers,
but it would help DBA understand what's actually happening around WAL
buffers.

Also, decreasing the WALWriteLock contention is obviously important
for DBA in terms of improving database performance.

Actually, that's the reason why I'm working on another statistics. :)
http://archives.postgresql.org/pgsql-hackers/2012-06/msg01489.php

Regards,
--
Satoshi Nagayasu <snaga(at)uptime(dot)jp>
Uptime Technologies, LLC. http://www.uptime.jp