| From: | Konrad Garus <konrad(dot)garus(at)gmail(dot)com> |
|---|---|
| To: | Cédric Villemain <cedric(dot)villemain(dot)debian(at)gmail(dot)com> |
| Cc: | pgsql-performance(at)postgresql(dot)org |
| Subject: | Re: shared_buffers advice |
| Date: | 2010-05-27 15:51:43 |
| Message-ID: | AANLkTimVO1B1XhbEcZmtkqrh-68Qd_fStwmJlwN-KlOp@mail.gmail.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
2010/5/27 Cédric Villemain <cedric(dot)villemain(dot)debian(at)gmail(dot)com>:
> well, that is the projection of file in memory. only projection, but
> the memory is still acquire. It is ok to rework this part and project
> something like 128MB and loop. (in fact the code is needed for 9.0
> because segment can be > 1GB, I didn't check what is the optimum
> projection size yet)
> So both yes at your questions :)
So when I map 12 GB, this process will consume 1 GB and the time
needed to browse through the whole 12 GB buffer?
--
Konrad Garus
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Craig Ringer | 2010-05-27 15:55:13 | Re: Query timing increased from 3s to 55s when used as function instead of select |
| Previous Message | Craig Ringer | 2010-05-27 15:33:06 | Re: Query timing increased from 3s to 55s when used as function instead of select |