| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Rafael Martinez <r(dot)m(dot)guerrero(at)usit(dot)uio(dot)no> |
| Cc: | Heikki Linnakangas <heikki(at)enterprisedb(dot)com>, pgsql-performance(at)postgresql(dot)org |
| Subject: | Re: Need to run CLUSTER to keep performance |
| Date: | 2007-11-08 15:33:00 |
| Message-ID: | 24775.1194535980@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
Rafael Martinez <r(dot)m(dot)guerrero(at)usit(dot)uio(dot)no> writes:
> Heikki Linnakangas wrote:
>> On a small table like that you could run VACUUM every few minutes
>> without much impact on performance. That should keep the table size in
>> check.
> Ok, we run VACUUM ANALYZE only one time a day, every night.
There's your problem.
Reading between the lines I gather that you think an update is "free"
in the sense of not creating a need for vacuum. It's not --- it's
exactly equivalent to an insert + a delete, and it leaves behind a
dead row that needs to be vacuumed. If you do a lot of updates, you
need to vacuum.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Chris Browne | 2007-11-08 15:47:44 | Re: Need to run CLUSTER to keep performance |
| Previous Message | Guillaume Smet | 2007-11-08 15:29:34 | Re: Estimation problem with a LIKE clause containing a / |