| From: | Kenneth Marshall <ktm(at)rice(dot)edu> |
|---|---|
| To: | Thomas Finneid <tfinneid(at)fcon(dot)no> |
| Cc: | Thomas Markus <t(dot)markus(at)proventis(dot)net>, pgsql-performance(at)postgresql(dot)org |
| Subject: | Re: caching indexes and pages? |
| Date: | 2009-01-22 22:03:33 |
| Message-ID: | 20090122220333.GG1961@it.is.rice.edu |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
On Thu, Jan 22, 2009 at 10:58:25PM +0100, Thomas Finneid wrote:
> Thomas Markus wrote:
>
>> try to reorganize your data with CLUSTER and create appropriate indixes
>> (dont forget to check statistics).
>
> One question. Assume I have clustered and new data has been added after
> that, according to the docs that data is added "outside" of the clustered
> data. What happens when I run cluster again? I would assume its smart and
> to only clusteres the new data, i.e. adding it to the already created
> clusters, as apporpriate, so the execution time would be a lot lower,
> right? or would it run through and recluster everything from scratch again?
>
> thomas
>
It reclusters again from scratch. You do get better performance on the
reads from the data that is already clustered.
Cheers,
Ken
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Ibrahim Harrani | 2009-01-22 22:29:57 | Re: postgresql 8.3 tps rate |
| Previous Message | Thomas Finneid | 2009-01-22 21:58:25 | Re: caching indexes and pages? |