Please chime in on MDADM stressing HDDs. Can it kill them?
by road hazard from LinuxQuestions.org on (#6E6VK)
I have an MDADM RAID 6 array with 20, 4TB drives plus 1 hot spare. I'm about to replace them all with 8TB Seagate drives (ST8000NM001A) and will be doubling the capacity. Over the years, I've added 4TB drives to my existing array (usually 2 at a time) and the rebuild took about 30 hours and I never ran into problems.
I think the easiest thing for me to do.... (so I can keep Plex up and running).... is to swap out one drive at a time, let the rebuild happen and after all 20 have been swapped out, grow the array to take full advantage of the new space. But.... that will mean a lot of non-stop HDD activity, 24 hours a day for 3 weeks. (Swap out the first drive, let it re-silver.... verify the array is good then swap the next one, rinse repeat.)
Has anyone attempted something like that before with that many drives? Will the heavy disk I/O lead to UREs or burn the drives out? The 4TB drives are WD Reds and they're in a SuperMicro case and current drive temps are about 85-90F.
I think the easiest thing for me to do.... (so I can keep Plex up and running).... is to swap out one drive at a time, let the rebuild happen and after all 20 have been swapped out, grow the array to take full advantage of the new space. But.... that will mean a lot of non-stop HDD activity, 24 hours a day for 3 weeks. (Swap out the first drive, let it re-silver.... verify the array is good then swap the next one, rinse repeat.)
Has anyone attempted something like that before with that many drives? Will the heavy disk I/O lead to UREs or burn the drives out? The 4TB drives are WD Reds and they're in a SuperMicro case and current drive temps are about 85-90F.