From: | Diogenes Caraballo <diogns(at)gmail(dot)com> |
---|---|
To: | Tena Sakai <tsakai(at)gallo(dot)ucsf(dot)edu> |
Cc: | pgsql-admin(at)postgresql(dot)org |
Subject: | Re: vacuuming and reindexing |
Date: | 2009-03-21 13:31:49 |
Message-ID: | c82353130903210631x71e1e8a0u65d2684ce2d48292@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-admin |
do you also update rows of that table?
because that action delete old one's and insert new one's..
if you are just inserting rows and not updating or deleting, i think is
correct your point.
Regards,
--
"Sin desafíos la vida es una rutina, una lenta agonía"
Diógenes Caraballo
On Fri, Mar 20, 2009 at 18:03, Tena Sakai <tsakai(at)gallo(dot)ucsf(dot)edu> wrote:
> Hi Everybody,
>
> I have a table that I add rows every evening.
> A rule I have for this table is that no row
> ever gets deleted. It has 600+ million rows.
>
> The last time I did it, it took nearly 4 hours
> to vacuum and 13 hours 40 minutes to reindex.
>
> My rudimental understanding of vacuuming is
> that when the rows get "deleted" it is only
> marked as "deleted" but the data is left alone
> until the act of vacuuming takes place. The
> vacuuming really gets rid of rows and pushes
> the data in such the way there is no "hole."
> And therefore after vacuuming, it is necessary
> to reindex (and analyze) the table.
>
> Is this a correct understanding?
>
> If it is (and as new rows get added new indexes
> are also built for the new rows), the fact that
> there is no deletion means there is no necessity
> for vacuuming this particular table?
>
> I appreciate any thoughts on this matter.
>
> Regards,
>
> Tena Sakai
> tsakai(at)gallo(dot)ucsf(dot)edu
>
>
>
>
From | Date | Subject | |
---|---|---|---|
Next Message | Tena Sakai | 2009-03-21 21:48:34 | Re: vacuuming and reindexing |
Previous Message | c k | 2009-03-21 08:37:40 | Accessing large objects |