From: | Paul Lambert <paul(dot)lambert(at)autoledgers(dot)com(dot)au> |
---|---|
To: | "pgsql-general(at)postgresql(dot)org" <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: Vacuuming |
Date: | 2007-05-08 02:58:51 |
Message-ID: | 463FE76B.6030009@autoledgers.com.au |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Tom Lane wrote:
>
> The only thing a vacuum would do for you there is set the commit hint
> bits on the newly-inserted rows. Which might be worth doing if you want
> to get the table into a totally "clean" state, but it's probably a bit
> excessive. SELECTs on the table will set the hint bits anyway as
> they visit not-yet-hinted rows, so it's really a matter of do you want
> to pay that overhead all at once or spread-out.
>
> What you *do* want to do in this situation is an ANALYZE.
>
> regards, tom lane
>
> ---------------------------(end of broadcast)---------------------------
> TIP 4: Have you searched our list archives?
>
> http://archives.postgresql.org/
>
>
Thanks Tom.
Should the ANALYZE be done before or after indexes are built? Or is that
irrelevant? Should I not even bother rebuilding indexes when I do these
loads?
Currently I:
1) Drop Indexes
2) Truncate and copy in new data
3) Vacuum - now changed to analyze.
4) Create indexes
I add steps one and four on the assumption that adding 40 million
records in one hit might get the indexes confused - but if they are
pretty stable I can remove these steps.
P.
--
Paul Lambert
Database Administrator
AutoLedgers
From | Date | Subject | |
---|---|---|---|
Next Message | Mariano Mara | 2007-05-08 03:06:58 | Re: Anyone know a good opensource CRM that actually installs with Posgtres? |
Previous Message | Rich Shepard | 2007-05-08 02:49:25 | Re: Date Math |