| From: | Renaud Thonnart <thonnart(at)amwdb(dot)u-strasbg(dot)fr> |
|---|---|
| To: | Einar Karttunen <ekarttun(at)cs(dot)Helsinki(dot)FI> |
| Cc: | pgsql-general(at)postgresql(dot)org |
| Subject: | Re: Time to insert |
| Date: | 2001-05-03 14:15:10 |
| Message-ID: | 3AF167EE.8778EC72@amwdb.u-strasbg.fr |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
Ok, I will try to do a VACUUM ANALYZE in my program.
I suppose it is better to do it too much than too less time....
Many thanks for your help Einar.
Renaud
Einar Karttunen wrote:
> On Thu, 3 May 2001, Renaud Thonnart wrote:
>
> > Thank you Einar and Colin for your help.
> >
> > I have some little qusetions more :
> >
> > - How do I interpret a VACUUM ANALYSE ?
> > - I'm going to load version 7.1 : what is the difference between 7.1 and
> > 7.1rc4 ?
> >
> 7.1rc4 is the fourth release candidate. 7.1 is the final release. So you
> should use 7.1.
>
> You just send "VACUUM ANALYZE" command to the backend. The vacuum analyze
> deletes rows, which are marked as unused and recreates the statistics
> about the table. Its use is essential (one time in 2-48h, or after large
> modifications), keeps the database running smoothly. You use vacuum
> analyze as any other command eg (using libpq++):
>
> PgDatabase data;
>
> data.ExecTuplesOk("MISC SQL QUERY")
>
> // when you need it
> data.ExecCommandOk("VACUUM ANALYZE");
>
> I think that it is better to do the vacuuming server side, if the
> application doesn't change a large amount of data.
>
> - Einar Karttunen
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andy Howarth | 2001-05-03 14:27:16 | Re: Problems compiling 7.1 with TCL support on Solaris 7 |
| Previous Message | Tom Lane | 2001-05-03 14:14:51 | Re: troubles with postgresql 7.0.2 |