Re: Hash join on int takes 8..114 seconds

From: tv(at)fuzzy(dot)cz
To: "Andrus" <kobruleht2(at)hot(dot)ee>
Cc: "Tomas Vondra" <tv(at)fuzzy(dot)cz>, "Scott Carey" <scott(at)richrelevance(dot)com>, "PFC" <lists(at)peufeu(dot)com>, pgsql-performance(at)postgresql(dot)org
Subject: Re: Hash join on int takes 8..114 seconds
Date: 2008-11-24 13:39:24
Message-ID: 60633.89.102.139.23.1227533964.squirrel@sq.gransy.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

>> Given the fact that the performance issues are caused by bloated tables
>> and / or slow I/O subsystem, moving to a similar system won't help I
>> guess.
>
> I have ran VACUUM FULL ANALYZE VERBOSE
> and set MAX_FSM_PAGES = 150000
>
> So there is no any bloat except pg_shdepend indexes which should not
> affect to query speed.

OK, what was the number of unused pointer items in the VACUUM output?

The query performance is still the same as when the tables were bloated?
What are the outputs of iostat/vmstat/dstat/top when running the query?

regards
Tomas

In response to

Responses

Browse pgsql-performance by date

  From Date Subject
Next Message Glyn Astill 2008-11-24 14:49:17 Re: Perc 3 DC
Previous Message Steve Clark 2008-11-24 13:23:54 Re: Perc 3 DC