Re: [PERFORM] perf problem with huge table

2010-02-11 Thread jesper
Hi all, i am trying to move my app from M$sql to PGsql, but i need a bit of help :) Except from all the other good advises about partitioning the dataset and such there is another aspect to keep in mind. When you have a large dataset and your queries become IO-bound the tuple density is going

Re: [PERFORM] moving pg_xlog -- yeah, it's worth it!

2010-02-11 Thread Bruce Momjian
Kevin Grittner wrote: Jesper Krogh jes...@krogh.cc wrote: Sorry if it is obvious.. but what filesystem/OS are you using and do you have BBU-writeback on the main data catalog also? Sorry for not providing more context. ATHENA:/var/pgsql/data # uname -a Linux ATHENA

[PERFORM] Dell PERC H700/H800

2010-02-11 Thread Matthew Wakeling
Just a heads up - apparently the more recent Dell RAID controllers will no longer recognise hard discs that weren't sold through Dell. http://www.channelregister.co.uk/2010/02/10/dell_perc_11th_gen_qualified_hdds_only/ As one of the comments points out, that kind of makes them no longer SATA

Re: [PERFORM] Re: 512,600ms query becomes 7500ms... but why? Postgres 8.3 query planner quirk?

2010-02-11 Thread Robert Haas
On Wed, Feb 10, 2010 at 8:52 PM, Bryce Nesbitt bry...@obviously.com wrote: If you guys succeed in making this class of query perform, you'll have beat out the professional consulting firm we hired, which was all but useless! The query is usually slow, but particular combinations of words seem

Re: [PERFORM] perf problem with huge table

2010-02-11 Thread Leo Mannhart
Dave Crooke wrote: On Wed, Feb 10, 2010 at 5:30 PM, Jon Lewison jlewis...@gmail.com mailto:jlewis...@gmail.com wrote: Just a nit, but Oracle implements MVCC. 90% of the databases out there do. Sorry, I spoke imprecisely. What I meant was the difference in how the rows are

Re: [PERFORM] moving pg_xlog -- yeah, it's worth it!

2010-02-11 Thread Scott Marlowe
On Thu, Feb 11, 2010 at 4:29 AM, Bruce Momjian br...@momjian.us wrote: Kevin Grittner wrote: Jesper Krogh jes...@krogh.cc wrote: Sorry if it is obvious.. but what filesystem/OS are you using and do you have BBU-writeback on the main data catalog also? Sorry for not providing more context.

Re: [PERFORM] Dell PERC H700/H800

2010-02-11 Thread Joshua D. Drake
On Thu, 2010-02-11 at 12:39 +, Matthew Wakeling wrote: Just a heads up - apparently the more recent Dell RAID controllers will no longer recognise hard discs that weren't sold through Dell. http://www.channelregister.co.uk/2010/02/10/dell_perc_11th_gen_qualified_hdds_only/ As one of

Re: [PERFORM] Dell PERC H700/H800

2010-02-11 Thread Scott Marlowe
On Thu, Feb 11, 2010 at 5:39 AM, Matthew Wakeling matt...@flymine.org wrote: Just a heads up - apparently the more recent Dell RAID controllers will no longer recognise hard discs that weren't sold through Dell.

Re: [PERFORM] moving pg_xlog -- yeah, it's worth it!

2010-02-11 Thread Kevin Grittner
Kevin Grittner kevin.gritt...@wicourts.gov wrote: Jesper Krogh jes...@krogh.cc wrote: Sorry if it is obvious.. but what filesystem/OS are you using and do you have BBU-writeback on the main data catalog also? File system is xfs noatime,nobarrier for all data; OS is on ext3. I *think*

Re: [PERFORM] moving pg_xlog -- yeah, it's worth it!

2010-02-11 Thread Alvaro Herrera
Kevin Grittner wrote: Another example of why I shouldn't trust my memory. Per the hardware tech: OS: /dev/sda is RAID1 - 2 x 2.5 15k SAS disk pg_xlog: /dev/sdb is RAID1 - 2 x 2.5 15k SAS disk These reside on a ServeRAID-MR10k controller with 256MB BB cache.

Re: [PERFORM] moving pg_xlog -- yeah, it's worth it!

2010-02-11 Thread Aidan Van Dyk
* Alvaro Herrera alvhe...@commandprompt.com [100211 12:58]: Hmm, so maybe the performance benefit is not from it being on a separate array, but from it being RAID1 instead of RAID5? Or the cumulative effects of: 1) Dedicated spindles/Raid1 2) More BBU cache available (I can't imagine the OS

Re: [PERFORM] moving pg_xlog -- yeah, it's worth it!

2010-02-11 Thread Kevin Grittner
Aidan Van Dyk ai...@highrise.ca wrote: Alvaro Herrera alvhe...@commandprompt.com wrote: Hmm, so maybe the performance benefit is not from it being on a separate array, but from it being RAID1 instead of RAID5? Or the cumulative effects of: 1) Dedicated spindles/Raid1 2) More BBU cache

Re: [PERFORM] Dell PERC H700/H800

2010-02-11 Thread Łukasz Jagiełło
2010/2/11 James Mansion ja...@mansionfamily.plus.com: Matthew Wakeling wrote: Just a heads up - apparently the more recent Dell RAID controllers will no longer recognise hard discs that weren't sold through Dell.

Re: [PERFORM] Dell PERC H700/H800

2010-02-11 Thread Scott Marlowe
On Thu, Feb 11, 2010 at 1:11 PM, James Mansion ja...@mansionfamily.plus.com wrote: Matthew Wakeling wrote: Just a heads up - apparently the more recent Dell RAID controllers will no longer recognise hard discs that weren't sold through Dell.

Re: [PERFORM] Dell PERC H700/H800

2010-02-11 Thread James Mansion
Matthew Wakeling wrote: Just a heads up - apparently the more recent Dell RAID controllers will no longer recognise hard discs that weren't sold through Dell. http://www.channelregister.co.uk/2010/02/10/dell_perc_11th_gen_qualified_hdds_only/ As one of the comments points out, that kind