[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: Lots of lost+found files ...


On Thu, Dec 13, 2001 at 12:28:37PM -0700, Andreas Dilger wrote:
> On Dec 06, 2001  18:23 +0000, Stephen C. Tweedie wrote:
> > On Thu, Dec 06, 2001 at 04:38:57PM -0000, Andy Bianchi wrote:
> > > Would I be right in thinking that if ext3 was hit with a data-corruption
> > > problem beyond it's control, i.e. it thought everything was okay but in
> > > reality something had got written wrong, that it would then be feasible 
> > > that I wouldn't notice it for a while since fsck wasn't been run?  Or is
> > > this theory garbage?
> > 
> > That's precisely correct, and this is why having a reliable fsck to
> > recover from unexpected corruption is important even for journaling
> > filesystems.
> And why disabling both the per-mount and time-based e2fscks is a bad idea.

Yes, it's a tradeoff people need to be aware of: disabling the checks
is good for availability but is a risk if you don't trust your

But the OS does give you some help.  Ext2 (and ext3, by derivation)
has the ability to mark the fs as having errors if it detects
corruption on the disk (by corruption it means genuine
inconsistencies, not just EIO sector read failures in harmless data
sectors), and if fsck sees such a flag on a filesystem, the full fsck
will be forced on the next boot even for ext3.


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]