| From: | Milan Zamazal <pdm(at)brailcom(dot)org> |
|---|---|
| To: | John R Pierce <pierce(at)hogranch(dot)com> |
| Cc: | Pavel Stehule <pavel(dot)stehule(at)gmail(dot)com>, pgsql-general(at)postgresql(dot)org |
| Subject: | Re: Large tables, ORDER BY and sequence/index scans |
| Date: | 2010-01-05 20:05:48 |
| Message-ID: | 87skak2u77.fsf@blackbird.nest.zamazal.org |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
>>>>> "JRP" == John R Pierce <pierce(at)hogranch(dot)com> writes:
>> effective_cache_size = 128MB
JRP> thats rather small unless your systme is very memory
JRP> constrained. assuming postgres is the primary disk IO consumer
JRP> on this ysstem, take a look at the cached' value on TOP or
JRP> whatever after its been running, thats a good first order
JRP> estimate of effective_cache_size.... this is often half or more
JRP> of your physical memory.
Indeed, increasing effective_cache_size helps, thanks.
Thank you all for the tips, I hope I understand the planner behavior for
my queries better now.
Regards,
Milan Zamazal
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Tim Uckun | 2010-01-05 22:14:37 | Re: timestams in the the pg_standby output |
| Previous Message | Guillaume Lelarge | 2010-01-05 20:00:34 | Re: reason for default PGSTAT_ACTIVITY_SIZE |