Re: Query Plans on Slaves

From: Tatsuo Ishii <ishii(at)postgresql(dot)org>
To: anthony(at)resolution(dot)com
Cc: pgsql-general(at)postgresql(dot)org
Subject: Re: Query Plans on Slaves
Date: 2014-11-10 02:50:02
Message-ID: 20141110.115002.949759977494642271.t-ishii@sraoss.co.jp
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-general

> Hi all!
>
> We're working on our architecture for our next set of systems, and we
> normally have a simple master/slave with wal shipping and hot_standby set
> up, with pgpool sitting in front of both to do load balancing.
>
> However, one piece I'm very confused about is the query plans on the slave
> server.
>
> Let's pretend we have a master server with:
> (2) E5-2660 CPU's
> (4) S3700 400GB SSD's
> 256GB of RAM
>
> And multiple slave servers that looks like:
> (1) E3-1290 CPU's
> (2) Intel 520 SSD's
> 32GB of RAM
>
> There's a significant difference between the master and slaves.

Can you show us the explain analyze result on both master and slave?

> Does the ANALYZE command run on each system, and work differently on each
> system, or would the slave servers use the query plans from the master
> machine?

ANALYZE can only be executed on master. Slave creates its own
plan. However since statistics should be same as master, I guess if
you get different plan, then postgresql.conf maybe different among
master and slave.

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

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message Anthony Presley 2014-11-10 03:17:14 Re: Query Plans on Slaves
Previous Message Anthony Presley 2014-11-10 02:15:02 Query Plans on Slaves