Linux Soft-Raid Fail

Soldato
Joined
25 Oct 2002
Posts
2,569
Location
Cambridge,York,Bristol
Hi All,

My mdadm raid just failed, here are the details. It includes 3 x Hitachi 1TB SATA drives in a Raid 5 arrangement. Suddenly, all 3 went from active sync to:

1 - spare
2 - active sync
3 - failed sync.

I managed to rescue the data by forcing 2&3 into an array and getting 99% of the data off. During this process the kernel was reporting a few problems getting data from (3), so I assumed this was the hard drive that caused the problem.

So now the data is saved I ran Hitachi's drive fitness check the drive I thought was the problem, but it reported (even with "full" "advanced test") the drive was fine. Same for the other two drives also.

Where do I go from here? Do I trust the drive checker and restart the array? Could the problem have been in the rest of the PC (memory corruption)?

Thanks for any pointers,
Sam
 
Back
Top Bottom