| From: | Simon Riggs <simon(at)2ndquadrant(dot)com> |
|---|---|
| To: | Stephen Frost <sfrost(at)snowman(dot)net> |
| Cc: | Luke Lonergan <LLonergan(at)greenplum(dot)com>, David Lang <dlang(at)invendra(dot)net>, Steve Oualline <soualline(at)stbernard(dot)com>, pgsql-performance(at)postgresql(dot)org |
| Subject: | Re: Database restore speed |
| Date: | 2005-12-03 12:38:48 |
| Message-ID: | 1133613528.2906.731.camel@localhost.localdomain |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
On Fri, 2005-12-02 at 15:18 -0500, Stephen Frost wrote:
> The other thought, of course, is that you could use PITR for your
> backups instead of pgdump...
Yes, it is much faster that way.
Over on -hackers a few optimizations of COPY have been discussed; one of
those is to optimize COPY when it is loading into a table created within
the same transaction as the COPY. This would allow pg_dumps to be
restored much faster, since no WAL need be written in this case.
I hope to work on this fairly soon.
Dumping/restoring data with pg_dump has wider uses than data protecting
backup.
Best Regards, Simon Riggs
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Jan Wieck | 2005-12-03 15:51:43 | Re: 15,000 tables - next step |
| Previous Message | David Lang | 2005-12-03 09:38:52 | Re: Database restore speed |