Re: More harrowing tales of TOAST growth - Mailing list pgsql-general

From Jeffrey Baker
Subject Re: More harrowing tales of TOAST growth
Date
Msg-id 20020505043034.GA19250@heat
Whole thread Raw
In response to More harrowing tales of TOAST growth  (Jeffrey Baker <jwbaker@acm.org>)
Responses Re: More harrowing tales of TOAST growth
List pgsql-general
On Sat, May 04, 2002 at 06:50:17PM -0700, Sean Chittenden wrote:
> > I ran this experiment for 10 minutes.  At time 0 the database size
> > was 2MB.  At 60 seconds the size was 22MB.  The size continued
> > upward:
> >
> > Time | Size
> > 0      2
> > 60     7
> > 120    60
> > 180    109
> > 240    181
> > 300    282
> > 360    284
> > 420    363
> > 480    442
> > 540    549
> > 600    557
>
> This may be outisde of the possibility of your environment, but have
> you tried a VACUUM FULL or VACUUM FREEZE ?  -sc

As I've said before, dump and reload is MUCH faster than vacuum
full, and they both require the same exclusive access.  Therefore
vacuum full isn't any better than my current situation.

-jwb

pgsql-general by date:

Previous
From: Feng-Cheng Chang
Date:
Subject: Fatal error after a system crash
Next
From: Hans Plum
Date:
Subject: Listing of Rules within psql?