From: | "HT" <htlevine(at)ebates(dot)com> |
---|---|
To: | pgsql-admin(at)postgresql(dot)org |
Subject: | Re: HELP? Postgres filling up disk |
Date: | 2002-12-19 18:13:45 |
Message-ID: | att28r$17hk$1@news.hub.org |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-admin |
Sorry, I've spent much time examining this in detail, so if I neglected the
* in the transcript I pasted you, it still doesn't matter. There is a .1
version (see my transcript below), BUT THAT'S IT . So my original
argument is still unanswered: there are still extra files on production
that aren't coming over in the standby : it's the .2 and .3 that I don't
want in the production DB. I have this pattern acros several large tables
and it's costing us disk space...
$ oid2name -d EBPRDS1 -t users
Oid of table users from database "EBPRDS1":
_______________________________
3828262123 = users
$ ls -l 3828262123*
-rw------- 1 postgres postgres 1073741824 Dec 18 23:08 3828262123
-rw------- 1 postgres postgres 558899200 Dec 18 23:16 3828262123.1
"Tom Lane" <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote in message
news:15552(dot)1040274982(at)sss(dot)pgh(dot)pa(dot)us(dot)(dot)(dot)
> HT Levine <htlevine(at)ebates(dot)com> writes:
> > For example, here is the oid2name dump of a table called users on
> > PRODUCTION:
> > $ oid2name -d EBPRD1 -t users
> > Oid of table users from database "EBPRD1":
> > _______________________________
> > 17260 = users
> > $ ls -l 17260*
> > -rw------- 1 postgres postgres 1073741824 Dec 18 17:48 17260
> > -rw------- 1 postgres postgres 1073741824 Dec 18 17:48 17260.1
> > -rw------- 1 postgres postgres 1073741824 Dec 18 17:48 17260.2
> > -rw------- 1 postgres postgres 165445632 Dec 18 17:51 17260.3
>
> > but the same table on the standby DB yields this:
>
> > $ oid2name -d EBPRDS1 -t users
> > Oid of table users from database "EBPRDS1":
> > _______________________________
> > 3828262123 = users
> > $ ls -l 3828262123
> > -rw------- 1 postgres postgres 1073741824 Dec 18 16:55 3828262123
>
> And? Your transcript does not prove that there's not 3828262123.1 and
> so forth ... given that 3828262123 is a full 1Gb, I'd be willing to bet
> lunch that 3828262123.1 exists too ...
>
> regards, tom lane
>
> ---------------------------(end of broadcast)---------------------------
> TIP 4: Don't 'kill -9' the postmaster
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2002-12-19 22:53:30 | Re: HELP? Postgres filling up disk |
Previous Message | Dan Langille | 2002-12-19 15:19:53 | Re: Sybase PowerDesigner Data Architect |