Re: Very big insert/join performance problem (bacula) - Mailing list pgsql-performance

From Kevin Grittner
Subject Re: Very big insert/join performance problem (bacula)
Date
Msg-id 4A5F422D0200002500028865@gw.wicourts.gov
Whole thread Raw
In response to Re: Very big insert/join performance problem (bacula)  (Marc Cousin <cousinmarc@gmail.com>)
Responses Re: Very big insert/join performance problem (bacula)
List pgsql-performance
Marc Cousin <cousinmarc@gmail.com> wrote:

> the hot parts of these 2 tables are extremely likely to be in the
> database or linux cache (buffer hit rate was 97% in the example
> provided). Moreover, the first two queries of the insert procedure
> fill the cache for us...

This would be why the optimizer does the best job estimating the
relative costs of various plans when you set the random_page_cost and
seq_page_cost very low.

-Kevin

pgsql-performance by date:

Previous
From: Scott Carey
Date:
Subject: Re: cluster index on a table
Next
From: "Kevin Grittner"
Date:
Subject: Re: cluster index on a table