| From: | Greg Smith <greg(at)2ndquadrant(dot)com> |
|---|---|
| To: | Kris Kewley <kris(dot)kewley(at)gmail(dot)com> |
| Cc: | Scott Marlowe <scott(dot)marlowe(at)gmail(dot)com>, Ashish Kumar Singh <ashishkumar(dot)singh(at)altair(dot)com>, "pgsql-performance(at)postgresql(dot)org" <pgsql-performance(at)postgresql(dot)org> |
| Subject: | Re: performance while importing a very large data set in to database |
| Date: | 2009-12-06 15:15:46 |
| Message-ID: | 4B1BCAA2.1010505@2ndquadrant.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
Kris Kewley wrote:
> Does postgres have the concept of "pinning" procs, functions, etc to
> cache.
>
No. Everything that's in PostgreSQL's cache gets a usage count attached
to is. When the buffer is used by something else, that count gets
incremented. And when new buffers need to be allocated, the process
that searches for them decrements usage counts until it find one with a
count of 0 that's then evicted. There's no way to pin things using this
scheme, the best you can do is try to access the data in advance and/or
regularly enough that its usage count never drops too far.
--
Greg Smith 2ndQuadrant Baltimore, MD
PostgreSQL Training, Services and Support
greg(at)2ndQuadrant(dot)com www.2ndQuadrant.com
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andres Freund | 2009-12-06 16:51:19 | Re: Large DB, several tuning questions: Index sizes, VACUUM, REINDEX, Autovacuum |
| Previous Message | Kris Kewley | 2009-12-06 14:15:59 | Re: performance while importing a very large data set in to database |