| From: | Tobias Brox <tobias(at)nordicbet(dot)com> |
|---|---|
| To: | Jeff Davis <pgsql(at)j-davis(dot)com> |
| Cc: | Tobias Brox <tobias(at)nordicbet(dot)com>, Paul Lathrop <plathrop(at)squaretrade(dot)com>, pgsql-performance(at)postgresql(dot)org |
| Subject: | Re: Defining performance. |
| Date: | 2006-12-01 01:07:54 |
| Message-ID: | 20061201010754.GA1448@oppetid.no |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
[Jeff Davis - Thu at 04:57:54PM -0800]
> > We're having the same issues, so we do the dumping and restoring every
> > now and then to be sure everything is properly cleaned up. With 8.1.
> >
>
> What's causing that? Is it index bloat?
>
> I would think a REINDEX would avoid having to dump/restore, right? A
> CLUSTER might also be necessary, depending on what kind of performance
> degradation you're experiencing.
>
> Am I missing something?
Just as with Paul Lathrops case, the performance degradation is
something perceived by the application developers. We haven't had time
to actually verify reliably that the performance is actually beeing
degraded, neither that the reload beeing done helps (after we resolved
the pending transaction issue, anyway), nor look into what the possible
reasons of this percieved degradation could be.
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Tobias Brox | 2006-12-01 01:15:54 | Re: Defining performance. |
| Previous Message | Jeff Davis | 2006-12-01 00:57:54 | Re: Defining performance. |