Skip site navigation (1) Skip section navigation (2)

Re: high user cpu, massive SELECTs, no io waiting problem

From: "Kevin Grittner" <Kevin(dot)Grittner(at)wicourts(dot)gov>
To: "Greg Smith" <greg(at)2ndquadrant(dot)com>, "Justin Pitts" <justinpitts(at)gmail(dot)com>
Cc: <pgsql-performance(at)postgresql(dot)org>, "Verteiler_A-Team" <a-team(at)turtle-entertainment(dot)de>, <bm(at)turtle-entertainment(dot)de>, "Felix Feinhals" <ff(at)turtle-entertainment(dot)de>, <tp(at)turtle-entertainment(dot)de>
Subject: Re: high user cpu, massive SELECTs, no io waiting problem
Date: 2011-02-16 16:08:54
Message-ID: 4D5BA236020000250003AB0E@gw.wicourts.gov (view raw or flat)
Thread:
Lists: pgsql-performance
Justin Pitts <justinpitts(at)gmail(dot)com> wrote: 
> I think adding
> 
> UNION ALL SELECT 'postgres version', version();
> 
> might be a good thing.
 
Good point.  Added.
 
> Greg Smith <greg(at)2ndquadrant(dot)com> wrote:
>> Kevin Grittner wrote:
>>>
>>> In fact, I wonder whether we shouldn't leave a couple items
>>> you've excluded, since they are sometimes germane to problems
>>> posted, like lc_collate and TimeZone.
>>
>> I pulled some of them out only because they're not really
>> postgresql.conf settings; lc_collate and lc_ctype for example are
>> set at initdb time.  Feel free to hack on that example if you
>> feel it could be improved, just be aware which of those things
>> are not really in the main config file when pondering if they
>> should be included.
 
Basically, the ones I could remember us needing to ask about on
multiple occasions, I put back -- provisionally.  If someone thinks
they're pointless, I won't worry about them being dropped again:
time zone, character encoding scheme, character set, and collation. 
I'm pretty sure I've seen us ask about all of those in trying to
sort out a problem.
 
I also tried the query on a newly installed HEAD build which had no
manual changes to the postgresql.conf file and found a few others
which seemed to me to be worth suppressing.
 
I took my shot -- anyone else is welcome to do so....  :-)
 
-Kevin

In response to

Responses

pgsql-performance by date

Next:From: Ross J. ReedstromDate: 2011-02-16 16:28:32
Subject: Re: Really really slow select count(*)
Previous:From: Scott MarloweDate: 2011-02-16 16:02:43
Subject: Re: high user cpu, massive SELECTs, no io waiting problem

Privacy Policy | About PostgreSQL
Copyright © 1996-2014 The PostgreSQL Global Development Group