| From: | Bruno Wolff III <bruno(at)wolff(dot)to> |
|---|---|
| To: | Joseph Bove <jbove(at)vetstar(dot)com> |
| Cc: | pgsql-performance(at)postgresql(dot)org |
| Subject: | Re: Inconsistent performance |
| Date: | 2003-09-15 21:42:28 |
| Message-ID: | 20030915214228.GA17642@wolff.to |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
On Mon, Sep 15, 2003 at 17:34:12 -0400,
Joseph Bove <jbove(at)vetstar(dot)com> wrote:
>
> I do a rather simple query: select count (*) from large-table where column
> = some value;
>
> About 80% of the time, the response time is sub-second. However, at 10% of
> the time, the response time is 5 - 10 seconds.
>
> This is nothing readily apparent at the system level that comes close to
> explaining the performance hits. CPU and memory usage (as measured by top)
> appear to be fine.
>
> Although there are certain tuning issues within the database itself, no
> documentation I have seen seems to indicate that tuning issues would lead
> to such inconsistent response time.
Looking at the output from explain analyze for the query would be useful.
It may be that there are a lot of rows that have the value in the problem
queries.
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Josh Berkus | 2003-09-15 21:42:59 | Re: restore time: sort_mem vs. checkpoing_segments |
| Previous Message | Stephan Szabo | 2003-09-15 21:34:49 | Re: Inconsistent performance |