Re: PostgreSQL 9.0 x64 bit pgbench TPC very low question?

From: tuanhoanganh <hatuan05(at)gmail(dot)com>
To: Kevin Grittner <Kevin(dot)Grittner(at)wicourts(dot)gov>
Cc: Gurjeet Singh <singh(dot)gurjeet(at)gmail(dot)com>, Magnus Hagander <magnus(at)hagander(dot)net>, pgsql-performance(at)postgresql(dot)org
Subject: Re: PostgreSQL 9.0 x64 bit pgbench TPC very low question?
Date: 2010-12-23 15:20:19
Message-ID: AANLkTikMO7Q4p5sTxRd=-Xw4tmspTUnGWZ7tgTHL6vRy@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

On Thu, Dec 23, 2010 at 9:37 PM, Kevin Grittner <Kevin(dot)Grittner(at)wicourts(dot)gov
> wrote:

> tuanhoanganh <hatuan05(at)gmail(dot)com> wrote:
>
> > Could you show me what parameter of pgbouncer.ini can do that. I
> > read pgbouncer and can not make pgbouncer open and keep 200
> > connect to postgres
>
> What makes you think that 200 connections to PostgreSQL will be a
> good idea? Perhaps you want a smaller number of connections from
> pgbouncer to PostgreSQL and a larger number from your application to
> pgbouncer?
>

If you read this thread, My app has ~ 20 exe file, each of exe create new
connect to postgesql and there are 10-30 user use my application.
My server running Windows 2008 R2. In this thread, postgresql on windows
create new connect very slow

"Starting a new connection in PG is relatively slow, especially so on
Windows, because it involves starting and setting up a new process for
each one.
Jeff Janes "

So I need a tool to open and keep 200 connect to postgres, my application
connect to this tool. And decrease time to connect to postgres ( because no
need to start new postgres process on windows)

>
> If you search the archives you can probably find at least 100 posts
> about how both throughput and response time degrade when you have
> more connections active then there are resources to use.
> (Saturation is often around twice the CPU core count plus the
> effective number of spindles, with caching reducing the latter.) It
> is quite often the case that a transaction will complete sooner if
> it is queued for later execution than if it is thrown into a mix
> where resources are saturated.
>
> -Kevin
>

In response to

Responses

Browse pgsql-performance by date

  From Date Subject
Next Message Desmond Coertzen 2010-12-23 16:34:31 Index on function that returns type with sub fields
Previous Message Kenneth Marshall 2010-12-23 14:46:28 Re: PostgreSQL 9.0 x64 bit pgbench TPC very low question?