> I had the same problem a week ago. If you use badblocks you have to 
> remember to specify the blocksize in order to get a useable result.
>
> If you use ext2 or ext3, a easier approach is probably to use e2fsck 
> -c. This will run badblocks and mark the badblocks as bad automaticly. 
> If you want you can use a couple of extra switches: e2fsck -cCfy (C = 
> show progress, f = force check, y = answer 'yes' to all questions). 
> All this is of course explained in "man e2fsck". Of course you should 
> be aware that important data may be lost for ever by running this 
> operation, so do a backup first.

Will take these options under consideration. The problem is that I was 
trying to take a backup of the computer using Symantec Ghost, and it 
failed after about 20%... thats what lead me to check for bad sectors.

> Since i did not have a TSL rescue disk i used a Knoppix disc to do 
> this from, however, you should probably do this from some kind of 
> rescue disk, or at least with the disk in question unmounted.

I had to do the same. I started badblocks in read-only mode, but had to 
cancel the test at one point since I needed the computer online again 
(it's still online, and I've not experienced any corrupted data, but 
then again the data in these sectors may not be used at all). It did 
spitt out a few sector numbers; I guess those were bad. Is it possible 
to check to see if these sectors contains "active" data (meaning 
non-deleted data)? If its not, there will be more secure to mark them as 
bad...

> Hope this can help. :)

Sure :)

> | Joacim Christiansen       +47 90 63 05 90       [EMAIL PROTECTED] |

/Christopher

_______________________________________________
tsl-discuss mailing list
[email protected]
http://lists.trustix.org/mailman/listinfo/tsl-discuss

Reply via email to