fsck

[Shawn T. Rutledge]

On Sun, Jun 01, 2036 at 05:37:09AM -0700, Kansir wrote:
> Ok Xmms locked up my system. No number-lock key light to be turned on or off.
> So I had to restart and of course hda2 came back as 0.9% non-contiguous and it
> was not unmounted cleanly. Now my question is. Is there any damage that might

If it had a good fsck, you're OK.  :-)  Especially if it auto-fixed
everything that it found.  If it makes you log in and fsck it by hand
and answer a bunch of questions, those are slightly more severe errors.
Still, I've never yet lost data on an ext2 partition due to these kinds
of problems (short of actual hardware problems like when a couple of 
drives died in that ill-fated RAID I put together a while back).

This is what all the fuss over journaling filesystems is about - they
make fscking obsolete.  As I understand it, it's sort of a state machine
or transactional solution... each change is made by writing to the 
journal; my guess is it would write just the disk block that needed 
changing.  If the whole changed block is not written when the machine
crashes, then that set of changes is lost but the original file remains
intact.  If the entire changed block did successfully get written, then
maybe it just changes a pointer or two and the new block stays in the
same place on the disk, but replaces the old block in the sequence of
blocks which make up the file.  But that would lead to fragmentation...
not sure what they do about that.  Anyway at every step, it's
impossible to get into an unknown state where the changes cannot either
be rolled back, or completed properly, with no user interaction.

Is anybody here using reiserFS or any of the others?

> not be seen by me? If so what do I do to fix this problem. because my system
> has just locked up again when opening a SO5.1 doc. Also is there any howto's
> on disk mirroring?

Sure, there's a RAID howto.  Separate ones for hardware and software
RAID I think.  (And I have nothing against RAIDs, just don't make the
mistake of cramming too many old-fashioned, hot-running SCSI drives into
too small a space without enough cooling or enough power supply "oomph"
to run them...)

-- 
  _______                   Shawn T. Rutledge / KB7PWD  ecloud@bigfoot.com
 (_  | |_)          http://www.bigfoot.com/~ecloud  kb7pwd@kb7pwd.ampr.org
 __) | | \________________________________________________________________
Get money for spare CPU cycles at http://www.ProcessTree.com/?sponsor=5903