Hard read errors in redundant array

[fawrell]

Hello everyone.

I'm wondering about one thing. When I have disk(s) in RAID 0 and some sectors aren't readable, the disk will produce hard read errors in SMART but it will not reallocate those sectors until some write action is done on those bad sectors. But what happends when hard read error occurs on a disk in redundant array on e.g. P600 controller? Will the controller be smart and after rebuilding those data from other disks will it tell the disk to reallocate those bad sectors, or will it be like in RAID 0?

Unfortunately (or luckily?) nothing like that has happened on our redundant arrays yet. Only on non redundant ones. So I would really like to know how is the controller handling this in rerundant array.

Thanks for your time!

[Uwe Zessin · Accepted solution]

A competent RAID system will attempt an operation called "Bad Block Repair". It will reconstruct the original data from the redundancy information and then attempt to replace the unreadable bad block.