(no subject)
Nov. 15th, 2005 11:51 pmLet it be known: having an openafs cache on ext3 is a bad idea. Especially when you combine it with software suspend 2. I'm wondering if any of my filesystem will be salvagable after this... sigh. Sucks. A lot.
e2fsck has been running for a while now... just waiting for it to say
"Your filesystem is fucked. Clear<y>?"
I really should be working on 212, too, but (clearly) not in the right mood at the moment.
(I suppose Windows will still work, hopefully, but still. Bah.)
e2fsck has been running for a while now... just waiting for it to say
"Your filesystem is fucked. Clear<y>?"
I really should be working on 212, too, but (clearly) not in the right mood at the moment.
(I suppose Windows will still work, hopefully, but still. Bah.)
(no subject)
Date: 2005-11-16 05:07 am (UTC)(no subject)
Date: 2005-11-16 05:39 am (UTC)(no subject)
Date: 2005-11-16 05:43 am (UTC)(no subject)
Date: 2005-11-18 01:43 am (UTC)-Deason
(no subject)
Date: 2005-11-18 04:01 am (UTC)That said, it's still known among openafs devs/people who use it extensively (I've talked to a few) that it's still a really bad idea to run your cache on a journaled filesystem, and I'm not the first person in the past few weeks to lose a system to afs mucking with things.
Your best bet is to either a) create an ext2 partition for afs's cache or b) create a loopback device on the local fs.
You may not have been burned yet, but you probably also haven't had a "bad situation" where the machine crashed (or just weird hacky things like hibernation). Just a warning so that you don't get burned in the future...