Re: metapartition information request

2006-07-10 Thread Matt Bernstein
On Jul 9 Matt Bernstein wrote: I've immediately noticed a bug in either imapd or squatter. Aaaargh.. actually the bug was in my head. The folder was created by webmail which uses a different imapd.conf (so as not to require SSL over the loopback interface) which I hadn't kept in sync with th

Re: metapartition information request

2006-07-09 Thread Matt Bernstein
On Jul 7 Ken Murchison wrote: Matt Bernstein wrote: Am I missing something? Sorry, I was. No, you can't have more than 1 metapartition per partition. OK, that's a shame. Is anyone else on this list as greedy as I am? :) For now, I've set up a metapartition just for the squat index files

Re: metapartition information request

2006-07-07 Thread Ken Murchison
Matt Bernstein wrote: On Jun 30 Ken Murchison wrote: Matt Bernstein wrote: I would like to use different metapartitions for different metadata. In particular, I'd like the squat indexes to be on their own volume--which won't need to be backed up--and all the small cyrus files on (yet) anot

Re: metapartition information request

2006-07-07 Thread Matt Bernstein
On Jun 30 Ken Murchison wrote: Matt Bernstein wrote: I would like to use different metapartitions for different metadata. In particular, I'd like the squat indexes to be on their own volume--which won't need to be backed up--and all the small cyrus files on (yet) another volume. Is this p

Re: metapartition information request

2006-06-30 Thread Ken Murchison
Matt Bernstein wrote: Hi, Having just upgraded FC4 -> FC5 (and thus Cyrus 2.2 -> 2.3), I'm starting to play with the new features in 2.3. I would like to use different metapartitions for different metadata. In particular, I'd like the squat indexes to be on their own volume--which won't nee

metapartition information request

2006-06-28 Thread Matt Bernstein
Hi, Having just upgraded FC4 -> FC5 (and thus Cyrus 2.2 -> 2.3), I'm starting to play with the new features in 2.3. I would like to use different metapartitions for different metadata. In particular, I'd like the squat indexes to be on their own volume--which won't need to be backed up--and