Skip site navigation (1) Skip section navigation (2)

Re: High CPU load on Postgres Server during Peak times!!!!

From: Karl Denninger <karl(at)denninger(dot)net>
To: Fernando Hevia <fhevia(at)ip-tel(dot)com(dot)ar>
Cc: "'Shiva Raman'" <raman(dot)shivag(at)gmail(dot)com>, pgsql-performance(at)postgresql(dot)org
Subject: Re: High CPU load on Postgres Server during Peak times!!!!
Date: 2009-09-23 20:52:01
Message-ID: (view raw, whole thread or download thread mbox)
Lists: pgsql-performance
Fernando Hevia wrote:
>>>> User Access
>>>> Total Number of Users is 500
>>>> Maximum number of Concurrent users will be 500 during peak time
>>>> Off Peak time the maximum number of concurrent user will be
>>>> around 150 to 200.
>>> A connection pooler like pgpool or pgbouncer would considerably reduce the
>>> burden on your system.
>> I am already using connection pooling in tomcat web server, so installing
> pgpool
>> will help enhancing the performance ?Any changes i have to do in my
> application to 
>> include pgpool? 
> There shouldn't be need for another pooling solution.
> Anyway, you probably dont want 1800 concurrent connections on your database
> server, nor even get near that number.
> Check the number of actual connections with: 
>   select count(*) from pg_stat_activity;
> A vmstat run during high loads could provide a hindsight to if the number of
> connections is straining your server.
> If the number of connections is high (say over 200-300), try reducing the
> pool size in Tomcat and see what happens.
> You possibly could do fine with something between 50 and 100 connections.
I can second this - I have an EXTREMELY busy forum system using pgpool
and during peak hours it runs very well within around 100 connections in

-- Karl

Attachment: karl.vcf
Description: text/x-vcard (124 bytes)

In response to

pgsql-performance by date

Next:From: Jared BeckDate: 2009-09-23 20:53:15
Subject: Slow query after upgrade to 8.4
Previous:From: Fernando HeviaDate: 2009-09-23 20:50:08
Subject: Re: High CPU load on Postgres Server during Peak times!!!!

Privacy Policy | About PostgreSQL
Copyright © 1996-2017 The PostgreSQL Global Development Group