From: | Bret Stern <bret_stern(at)machinemanagement(dot)com> |
---|---|
To: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: Backups |
Date: | 2012-03-15 17:04:42 |
Message-ID: | 1331831082.7259.2.camel@fedora13 |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Perhaps a RAM DISK could be considered in the equation
On Thu, 2012-03-15 at 16:30 +0100, Albe Laurenz wrote:
> Richard Harley wrote:
> > Very simple question - does pg_dump/dumpall hit the server in terms
> of database performance? We
> > currently do nightly backups and I want to move to hourly backups but
> not at the expense of hogging
> > all the resources for 5 mins.
> >
> > Pg_dumpall is currently producing a 1GB file - that's the combined
> size of around 30 databases and it
> > takes around 5 mins to run.
>
> pg_dump will be a performance hit, because it consumes disk I/O
> capacity.
> You could measure how the system is affected by your current backup.
>
> On the other hand, instead of doing an hourly pg_dump, it might be
> preferable to do a filesystem backup and PITR. That way you have to
> do a backup only once a day (or week, depends how much traffic you have
> and how fast you have to restore) and can still recover to an
> arbitrary point in time.
>
> Yours,
> Laurenz Albe
>
From | Date | Subject | |
---|---|---|---|
Next Message | Lonni J Friedman | 2012-03-15 19:22:38 | how to measure wal_buffer usage |
Previous Message | Rob Sargent | 2012-03-15 16:22:29 | Re: Did xpath_table get dropped. |