Hello!
On Mon, Feb 18, 2002 at 03:12:34PM +0100, Jens Benecke wrote:
> > > > (have you tried 2.5.3/2.5.4-pre1 kernels there?)
> > > No. I haven't tried 2.5.x. kernels yet and I'm not about to.
> > Just making sure. Error that can cause these items in wrong order
> > errors was fixed recently, but before it was believed to only cause
> > problems on 2.5, so now we know it can happen on 2.4, too.
> I'd be happy to backtrack this if there is a way. (Perhaps I should have
> saved the metadata before the fsck... well, too late).
This is easily reproducable on 2.5, so there is no point to do it on 2.4.
They 2.4 and 2.5 share most of the code, anyway.
> > > See other post, but I cannot reproduce all of this. I really don't
> > > know what went wrong.
> > Probably you used 2.4. kernel without the fix, and then went to
> > 2.4.18-rc1 with a lot of fixes. And these fixes noticed problems.
> I went from 2.4.15pre1 to 2.4.18pre3, which crashed within a couple
> days, so I tried 2.4.18pre7, which also crashed, then I switched some of
> the grsecurity things off and tried pre9, which went well for a few days
> but suddenly didn't let me log on any more, then I tried rc1 with even
> less grsecurity stuff enabled and then all hell broke loose.
> Back to pre9, everything normal, fsck, back to rc1, everything normal -
> for now.
Basically, I think if you read some errors were noticed in some filesystem,
and then fixed and you plan to upgrade to that kernel,
it is better to run fsck first, just in case. So that you are
sure you was not bitten by previous errors (and you also have
perfectly valid reason to argue that new fixes broke something, if new
code breaks and there were no errors on the filesystem before new code was run).
> > > A basic problem I have with ReiserFS is that the journaling makes
> > > you forget about hard disk errors until you get lots of "permission
> > > denied"s, at which time it is usually quite late to do something.
> > Journal is in no direct relation to those "permission denied"s, that
> > data is not from journal.
> Sorry, what I meant is that the journaling works so "well" that you
> don't notice there is something wrong with your disk any more, which
> perhaps the journaling did NOT fix - until you spot a file that isn't
> accessible any more.
No. Journaling does not wirk this way. If you have HDD errors, you'd
find their traces in system log pretty quickly.
> So you can have a corrupted file system without noticing anything
> (because the system came up without errors after the power failure). On
Yes, lack of fsck run on system startu may cover some stuff,
but after all journaling filesystem is supposed to have at least consistent
metadata, so it should not need fsck. Bugs is the other issue, though ;)
> a non-journaling FS, once the OS spots something is wrong it checks the
> _whole_ disk, and so also finds errors that were not related to the
> crash.
But you waste tons of time, that's why people are converting to journaling
filesystems.
> > If you need such a feature, you can easily implement it in your
> > initscripts.
> How do I find out the mount count of a ReiserFS partition?
Sigh. No easy way I can see. But your request is heard and next version of
reiserfsdebug will print it.
Bye,
Oleg
