[lug] Bad Sectors

[Zan Lynx]

Gordon Golding wrote:
>   I have a disk drive (used for data)  that started giving "messages".
> 
> It happened right after the system drive filled up and crashed.
> 
> SMART test says:
> 
> Powered on 978 days
> Current Pending Sector Count   10 sectors   FAILING  Old-age
> Uncorrectable Sector Count        7 sectors   N/A         Old-age
> 
> Questions:
> 1)  Could the fact the the system drive crashed in the middle of a BIG 
> tar on this drive have something to do with it?
> Should I try anything before believing this?
> 
> 2)  This is a WD 320 - I thought these drives had big bad sector tables.
> Is it time to just junk this unit?

It appears to be about three years old. That is about the right age for 
common drive failures.

However, if by "crash" you meant that the power went out, or someone 
reset the power to recover the system, the drive could have failed to 
write the sectors and is detecting them as "bad", but those bad sectors 
could be remapped by rewriting them.

I think the bad block test program in read/write/read non-destructive 
write mode can do it, or if you want to take a risk you can try using 
"dd" on the block device to write zeros to the bad sectors if you know 
where they are and know how to do the math to convert the dmesg sector 
logs to dd byte/block offsets.

-- 
Zan Lynx
zlynx at acm.org

"Knowledge is Power.  Power Corrupts.  Study Hard.  Be Evil."