People,
I must have missed something.
Disk drives have had for some years the logic to detect and remap
detected bad sectors on the fly. Some drives did this silently while
others <may> have had additional functionality to report a developing
problem. At any rate, the disks have already dealt with the problem
(reported or not), so remapping a sector (that is already remapped)
makes not one whit of sense. As has already been said, if the disk
has run out of spare sectors, the disk is in serious trouble
already and probably scraping itself out of existence.
Given this, why insert code into a filesystem to deal with errors
that are already remedied? Perhaps the device driver might stimulate
an alarm or at least an entry in the syslog so the admin would be aware
of a developing failure. That is not a filesystem's task. What should
the filesystem do? Handle errors from a device that doesn't remap?
How much responsibility does the filesystem have with regard to hardware
reliability?
-- Reid
Walter Reid Fletcher, WB7CJO
Department of Geology and Geophysics P. O. Box 3006
University of Wyoming 1-307-766-6227
Laramie, WY 82071 Internet: Fletcher @ UWyo.Edu
Are we all roadkill on the information highway? - Jeff Greenfield
|