Re: High CPU usage after partitioning

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Andrew Dunstan <andrew(at)dunslane(dot)net>
Cc: rudi <rudolone(at)gmail(dot)com>, pgsql-performance(at)postgresql(dot)org
Subject: Re: High CPU usage after partitioning
Date: 2013-01-22 18:38:28
Message-ID: 24018.1358879908@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

Andrew Dunstan <andrew(at)dunslane(dot)net> writes:
> On 01/22/2013 09:21 AM, rudi wrote:
>> The query is pretty simple and standard, the behaviour (and the plan)
>> is totally different when it comes to a partitioned table.
>>
>> Partioned table query => explain analyze SELECT "sb_logs".* FROM
>> "sb_logs" WHERE (device_id = 901 AND date_taken = (SELECT
>> MAX(date_taken) FROM sb_logs WHERE device_id = 901));

> And there you have it. Constraint exclusion does not work in cases like
> this. It only works with static expressions (such as a literal date in
> this case).

This isn't about constraint exclusion I think. The main problem is in
the sub-select: 9.0 isn't able to index-optimize a MAX() across a
partitioned table, for lack of MergeAppend, so you end up scanning lots
of rows there. 9.1 or 9.2 should be better.

regards, tom lane

In response to

Browse pgsql-performance by date

  From Date Subject
Next Message Виктор Егоров 2013-01-22 20:57:50 Effect of the WindowAgg on the Nested Loop
Previous Message Igor Neyman 2013-01-22 15:42:46 Re: High CPU usage after partitioning