| From: | "Kevin Grittner" <Kevin(dot)Grittner(at)wicourts(dot)gov> |
|---|---|
| To: | "Hannu Krosing" <hannu(at)2ndquadrant(dot)com>, "Tom Lane" <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
| Cc: | "Robert Haas" <robertmhaas(at)gmail(dot)com>, "Peter Hussey" <peter(at)labkey(dot)com>, "pgsql-performance" <pgsql-performance(at)postgresql(dot)org> |
| Subject: | Re: Questions on query planner, join types, and work_mem |
| Date: | 2010-08-04 20:42:49 |
| Message-ID: | 4C598A7902000025000341FF@gw.wicourts.gov |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
Hannu Krosing <hannu(at)2ndquadrant(dot)com> wrote:
> This seems to indicate about 1 ms for moving pages over
> user/system boundary. (Intel Core2 Duo T7500 @ 2.20GHz, Ubuntu
> 9.10, 4GB RAM)
Using Greg's test script on a box with two cores like this:
Intel(R) Pentium(R) D CPU 3.40GHz
Linux kgrittn-desktop 2.6.31-22-generic #60-Ubuntu SMP Thu May 27
00:22:23 UTC 2010 i686 GNU/Linux
Dividing the run time by accumulated buffers_backend, it comes to
less than 0.2 ms per dirty buffer flushed. If I get a few spare
ticks I'll try again while checking what vmstat and oprofile say
about how much of that went to things besides the transfer from
shared buffers to the OS. I mean, it's possible I was waiting on
actual disk I/O at some point.
-Kevin
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Kevin Grittner | 2010-08-04 20:46:05 | Re: Questions on query planner, join types, and work_mem |
| Previous Message | Greg Smith | 2010-08-04 20:40:14 | Re: Questions on query planner, join types, and work_mem |