| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> | 
|---|---|
| To: | yamwak(at)yahoo(dot)com | 
| Cc: | pgsql-general(at)postgresql(dot)org | 
| Subject: | Re: Horrible pg_restore performance, please help | 
| Date: | 2003-07-20 05:16:10 | 
| Message-ID: | 20178.1058678170@sss.pgh.pa.us | 
| Views: | Whole Thread | Raw Message | Download mbox | Resend email | 
| Thread: | |
| Lists: | pgsql-general | 
brice <yamwak(at)yahoo(dot)com> writes:
> We're running Postgres 7.1.3 on a Redhat Linux system with 800 MHz CPU,
> 1GB ram, 1.5 GB Swap. This is obviously not a screamer of a box, but
> its not terrible. 
> We're trying to import a 32 MB dump file and its taking forever. We've
> tried multiple cfg changes, but things don't seem to speed up, or
> rather ever finish. After about 10 minutes, the cpu is 100% consumed by
> system, with pg_restore being the proc on top. pg_restore consumes
> about 1200 MB of memory in total.
pg_restore, not the backend?
Some digging in the CVS logs reveals past fixes for memory leaks and
other problems in pg_restore, particularly when dealing with lots of
large objects.  If you have LOs in your database then an update to 7.2
or 7.3 should improve life.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Tom Lane | 2003-07-20 05:24:10 | Re: cidr & inet types | 
| Previous Message | Joe Conway | 2003-07-20 05:13:12 | Re: Detoasting and memory usage |