From: | Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us> |
---|---|
To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
Cc: | Justin Clift <justin(at)postgresql(dot)org>, pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: Brain dump: btree collapsing |
Date: | 2003-02-13 03:58:23 |
Message-ID: | 200302130358.h1D3wNj24688@candle.pha.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Tom Lane wrote:
> Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us> writes:
> > It would be nice if VACUUM FULL would be able to compress the actual
> > index file and return unused space to the operating system. REINDEX
> > does this, but I was thinking of something a little lighter that could
> > be done automatically as part of VACUUM FULL.
>
> But indexes tend to be very dependent on physical layout. You can't
> just shove stuff around without thinking about the consequences.
> Tables (heaps) are *much* more forgiving about that.
>
> My feeling is that what we need to fix now is index bloat during normal
> operation. If you want the indexes to actually *shrink*, that's a job
> for REINDEX. Perhaps someday we can improve on that --- but let's not
> blur our focus on the immediate problem.
My point is only that while we need VACUUM and VACUUM FULL to match all
heap needs, we need a VACUUM FULL capability for indexes too. REINDEX
may be that capability, but it would be nice if we could compress out
some index space during VACUUM FULL.
--
Bruce Momjian | http://candle.pha.pa.us
pgman(at)candle(dot)pha(dot)pa(dot)us | (610) 359-1001
+ If your life is a hard drive, | 13 Roberts Road
+ Christ can be your backup. | Newtown Square, Pennsylvania 19073
From | Date | Subject | |
---|---|---|---|
Next Message | Bruce Momjian | 2003-02-13 04:08:16 | Re: psql and readline |
Previous Message | Greg Stark | 2003-02-13 03:41:55 | Re: log_duration |