regular fsck runs are too disturbing

mike corn mikecorn at
Thu Oct 11 05:20:41 UTC 2007

For this to be true, you need another assumption: All hardware is 
absolutely reliable which just is not the case.

Windows runs on the same "potentially flakey" hardware that Linux does, 
and it doesn't routinely perform a chkdsk.  Most people are quite happy 
with this and only need to chkdsk when something goes wrong and they 
suspect filesystem damage.  The argument about random hardware 
corruption does not hold up in the face of this evidence.

Yes, but...

Running fsck unconditionally every N boots is a crude solution. It 
ignores the fact that some systems have done millions of file operations 
and others have done a few thousand. It ignores the availability of 
hardware health/status information available from modern disks 
(different for every make/model?).

So my question is: can it be made IO volume dependent? Can it make use 
of hardware status information (i.e. run fsck unconditionally if there 
were more than normal rate of soft errors (ECC corrections) or bad tracks)?

More information about the Ubuntu-devel-discuss mailing list