On Tue, Nov 15, 2005 at 01:53:27PM +0800, ITS ONT Alcazar, Jose Aguedo C wrote:
> Thanks Tom! I think we did it. Below are the thing's i've inputed. Also, i
> have change the attributes and ownership of the xlog file as to the same
> with the previous file. Probably, they (sysadmins) have restored the backup
> xlog file to the original folder. That's why it was changed the attributes
> and ownership. Do you think this will resolve the issue? BTW, can you
> suggest on the clean-up thing? I think I wont recommend the one they are
> doing right now, deleting the files in pxlog. whats the best way they can do
> the clean-up? which one do they need to be "deleted"? Please advise. Thanks!
Actually, given that you are nearly full on that disk, you probably
need more disk. You ought to be able to trim some things out of
/var/log. But you can't delete anything from the PostgreSQL area
(/var/lib/pgsql/*, it looks like). There's no "clean up" you can do
there -- your data is there.
You _may_ need a vacuum full -- if you have a lot of dead tuples, you
might be able to recover some disk space. It'll lock all the tables,
so you'll need to be able to take an outage to do this. In the psql
command monitor (psql -U postgres [yourdbname]) you can issue a
VACUUM FULL VERBOSE and recover available space (and find out how
much you're recovering -- that's what the VERBOSE does). Note that
this may take a _long_ time to complete (depending how big oyur
database is), and (as I noted) it will cause your application not to
work while it's going on.
A
--
Andrew Sullivan | ajs@crankycanuck.ca
"The year's penultimate month" is not in truth a good way of saying
November.
--H.W. Fowler