1

I understand that when a disk fails in a RAID configuration, whatever the RAID level you're using (1, 5, 6, or 10), you need to replace the failed disk and the array rebuilds. Now with large capacity, consumer grade disks, the probability of encountering a URE during rebuild is quite high.
What I don't understand is what happens then? Does that mean the whole array is rendered immediately useless and needs to be restored from backup (after you replace the URE-causing disk, I guess), or can it still work in degraded mode (if the configuration allows it)?

2

In general, the array rebuild will stop as soon as it hits a URE, that's why it's a concern.

The RAID controller is trying to rebuild the array so that it's data is 100% accurate to before the rebuild was needed. If it can't do that (due to a URE, etc.) then the rebuild fails. At that point you need to test each disk, and if they test OK (and you trust them), then restore from backup.

Take RAID5 for example. It can tolerate 1 drive failure. So you have a drive fail, and you start the rebuild. A URE happens, so either the drive it's reading from is bad, or because the controller thinks it is, at that point the drive gets marked off-line. So that means you've now have two drives failed in a RAID that can only handle 1 drive failure at a time.

Perhaps also check out this related SU question: Why URE fails raid rebuild and "renders RAID 5 unusable"

  • Thanks, that's clear now. I had the question you mention but it doesn't answer the specifics of what happens in case of rebuild URE. – Antoine Feb 11 '15 at 22:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.