On Wed, Oct 2, 2013 at 2:55 AM, Robin Hill <robin@xxxxxxxxxxxxxxx> wrote: > On Tue Oct 01, 2013 at 03:42:39 -0500, Mark Keisler wrote: > >> Is scrub (sync) needed on RAID10? I've noticed in mdadm monitor logs that >> it seems to re-allocate data destined for bad blocks on the fly anyway. >> Maybe I don't quite understand what the scrub does, though :) >> I ran it for the first time on my RAID last night anyway to make sure all >> way well since I am failing out a dying disk for replacement today. > > "Scrubbing" will read all the data on all the disks in the array, which > ensures there's no read errors (as well as correcting any parity/mirror > mismatches). Any found will then get rewritten (which may cause a bad > block re-allocation in the background) while the parity/mirrored data is > still available. If there's too many in a short time then the disk will > be failed. > > Otherwise you'll only find out about read errors if the data happens to > be read in normal use, or when you're trying to recover a failed disk > (and unless you have multiple levels of redundancy, you're then > stuffed). > > I'd recommend running it regularly on any software RAID setup (well, any > with redundancy anyway). > > Cheers, > Robin Thanks for the explanation, I like to avoid getting stuffed :) -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
