[pgpool-general: 3995] Re: accept() scalability issues

Tatsuo Ishii ishii at postgresql.org
Fri Aug 21 07:50:24 JST 2015


> “Somewhat” would be the understatement of the year in our scenario:
> 
> http://imgur.com/a/EzTB3 <http://imgur.com/a/EzTB3>
> 
> Upgraded from pgpool2 3.3.6 to 3.3.7 including your semaphore patch. Switchover is clearly visible in the graphs.

I was hoping to see imporvements in pgbench -C or similar use case
(for example ab), which shows the connection overhead from client to
pgpool-II.

> If there are no adverse effects/broken features (idle timeouts?) due to your patch I highly recommend integrating it both in 3.3 and 3.4, at least as compiletime or config option 

Yeah, I noticed that with my patch child_life_time will not work
anymore. Let me see if I can remove the downside in the patch. If I
could do it, I think it's resonable to back pach to older branches.

Best regards,
--
Tatsuo Ishii
SRA OSS, Inc. Japan
English: http://www.sraoss.co.jp/index_en.php
Japanese:http://www.sraoss.co.jp


More information about the pgpool-general mailing list