| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Franco Bruno Borghesi <franco(at)akyasociados(dot)com(dot)ar> |
| Cc: | Josh Berkus <josh(at)agliodbs(dot)com>, Daniel Migowski <postgresql(at)Mig-O(dot)de>, pgsql-performance(at)postgresql(dot)org |
| Subject: | Re: Mapping a database completly into Memory |
| Date: | 2003-07-28 16:25:57 |
| Message-ID: | 10185.1059409557@sss.pgh.pa.us |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
Franco Bruno Borghesi <franco(at)akyasociados(dot)com(dot)ar> writes:
> wouldn't also increasing shared_buffers to 64 or 128 MB be a good
> performance improvement? This way, pages belonging to heavily used
> indexes would be already cached by the database itself.
Not necessarily. The trouble with large shared_buffers settings is you
end up with lots of pages being doubly cached (both in PG's buffers and
in the kernel's disk cache), thus wasting RAM. If we had a portable way
of preventing the kernel from caching the same page, it would make more
sense to run with large shared_buffers.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Josh Berkus | 2003-07-28 16:50:36 | Re: Mapping a database completly into Memory |
| Previous Message | Justin Long | 2003-07-28 16:18:55 | Optimization |