Jeff Frost wrote:
Tena Sakai wrote:
> If you want to restore all the databases that were in this
> postgresql installation at the time of the backup, then the
> best thing to do is drop all those DBs before attempting the
> restore.
Yes, the database in question is built and updated continuously
from a several sources. There was a massive power failure,
a series of them, and things got to be a very inconsistent
state and therefore we need to go back to a reliable, trustworthy
backup and then rebuild from there.
What I gather, from your comments, all I have to do would to issue
a psql command:
drop database <dbname>
then repeat what I did from shell prompt, ie.,
zcat <compressed_file> | psql postgres > restore.out 2 > restore.err
Would you mind confirming if I am understanding you correctly?
Yes, based on the information you've given us, you should be able to restore the entire database (and any other databases that were in the cluster) by first dropping those databases and then issuing the above command.
BTW, if you find yourself with an older version of postgresql, this could be a good opportunity to upgrade. I'm not sure if you mentioned what version you were using in your original post.
I should also note that you could rename the database instead of dropping it outright, to make sure your restore is effective before dropping it.
--
Jeff Frost, Owner <jeff@frostconsultingllc.com>
Frost Consulting, LLC http://www.frostconsultingllc.com/
Phone: 916-647-6411 FAX: 916-405-4032