From: | Vivek Khera <khera(at)kcilink(dot)com> |
---|---|
To: | pgsql-performance(at)postgresql(dot)org |
Subject: | Re: pg_restore taking 4 hours! |
Date: | 2004-12-13 16:37:28 |
Message-ID: | x73byab1dj.fsf@yertle.int.kciLink.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general pgsql-performance |
>>>>> "RC" == Rodrigo Carvalhaes <grupos(at)carvalhaes(dot)net> writes:
RC> Hi!
RC> I am using PostgreSQL with a proprietary ERP software in Brazil. The
RC> database have around 1.600 tables (each one with +/- 50 columns).
RC> My problem now is the time that takes to restore a dump. My customer
RC> database have arount 500mb (on the disk, not the dump file) and I am
RC> making the dump with pg_dump -Fc, my dumped file have 30mb. To make
RC> the dump, it's taking +/- 1,5 hours BUT to restore (using pg_restore )
RC> it it takes 4 - 5 hours!!!
I regularly dump a db that is compressed at over 2Gb. Last time I did
a restore on the production box it took about 3 hours. Restoring it
into a development box with a SATA RAID0 config takes like 7 hours or
so.
The biggest improvement in speed to restore time I have discovered is
to increase the checkpoint segments. I bump mine to about 50. And
moving the pg_xlog to a separate physical disk helps a lot there, too.
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Vivek Khera, Ph.D. Khera Communications, Inc.
Internet: khera(at)kciLink(dot)com Rockville, MD +1-301-869-4449 x806
AIM: vivekkhera Y!: vivek_khera http://www.khera.org/~vivek/
From | Date | Subject | |
---|---|---|---|
Next Message | Bruno Wolff III | 2004-12-13 16:37:45 | Re: table with sort_key without gaps |
Previous Message | Hans-Michael Stahl | 2004-12-13 16:36:16 | "Complex" data types like BYTEA in embedded SQL with ecpg |
From | Date | Subject | |
---|---|---|---|
Next Message | Vivek Khera | 2004-12-13 16:47:34 | Re: Alternatives to Dell? |
Previous Message | Vivek Khera | 2004-12-13 16:33:47 | Re: Hardware purchase question |