Re: Poor performance o - Mailing list pgsql-performance

From Tom Lane
Subject Re: Poor performance o
Date
Msg-id 10012.1143001903@sss.pgh.pa.us
Whole thread Raw
In response to Re: Poor performance o  ("Craig A. James" <cjames@modgraph-usa.com>)
List pgsql-performance
"Craig A. James" <cjames@modgraph-usa.com> writes:
> Tom Lane wrote:
>> Given the sizes of the tables involved, you'd likely have to boost up
>> work_mem before the planner would consider a hash join.  What nondefault
>> configuration settings do you have, anyway?

> shared_buffers = 20000
> work_mem = 32768
> effective_cache_size = 300000

So for a 6M-row table, 32M work_mem would allow ... um ... 5 bytes per
row.  It's not happening :-(

Try boosting work_mem by a factor of 100 and seeing whether a hash-based
join actually wins or not.  If so, we can discuss where the sane setting
really falls, if not there's no point.

            regards, tom lane

pgsql-performance by date:

Previous
From: Jeff Frost
Date:
Subject: motherboard recommendations
Next
From: Guillaume Cottenceau
Date:
Subject: Re: planner with index scan cost way off actual cost,