[pgpool-general: 1447] Re: Performance SELECT

[Francesco Di Girolamo]

My bottle neck  is about the " Large queries", the queries with more
100.000 records table. The time of decadence is enormous.
For a 1104 records result 8 seconds in direct connection and 1 minute 5 s
in pgpool streaming mode.


2013/2/26 Tatsuo Ishii <ishii at postgresql.org>

> You could always use standard performance tools such as sysstat,
> vmstat etc. to know which is the bottle neck.
> --
> Tatsuo Ishii
> SRA OSS, Inc. Japan
> English: http://www.sraoss.co.jp/index_en.php
> Japanese: http://www.sraoss.co.jp
>
> > Instead when i execute simultaneous async queries (select) with AJAX, i
> > notice my server use a lot of processes. These processes use resources.
> Is
> > only the hardware and the net its bottleneck ?
> > thk a lot
> > Franz
> >
> >
> > 2013/2/26 Tatsuo Ishii <ishii at postgresql.org>
> >
> >> > Is there a guide about performance setting by hardware ? I have an
> hard
> >> > decrement in my select queries. I used streaming replication on two
> >> windows
> >> > server with PG 8.3. Sorry but client's requisites do not permit to
> >> migrate
> >> > to Linux. :-(
> >>
> >> There's no performance documentation yet. However I think the major
> >> bottle neck is network, including netowrk hardware and TCP/IP stack.
> >>
> >> Also too trivial SELECT (like pgbench -S) will mask the effect of
> >> pgpool performance boosting in load balancing.
> >> --
> >> Tatsuo Ishii
> >> SRA OSS, Inc. Japan
> >> English: http://www.sraoss.co.jp/index_en.php
> >> Japanese: http://www.sraoss.co.jp
> >>
> >
> >
> >
> > --
> > *Franci anni'80*
> > ♫
> > (\ /)
> > ( . .) ♥●•٠·˙
> > c(")(")
>



-- 
*Franci anni'80*
♫
(\ /)
( . .) ♥●•٠·˙
c(")(")
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.sraoss.jp/pipermail/pgpool-general/attachments/20130228/5f4dc94a/attachment.html>