February 3, 201511 yr I am eager to use this 4 TB drive in a 2nd unraid mahcine as due to a recent hardware failure as my data is no longer backed up. The first prelcear reached 40% post-read when a blizzard took out our power for an extended period. The second preclear reached 98% post-read when I had to force the server down due to a mover mishap. This preclear will finish tomorrow evening (or so I hope). Given that the first 2 preclears reached post-read is it safe to trust the reports from the final preclear only... or should i really run 2 more full preclears? This is an HGST 4 TB drive. Thanks crew
February 3, 201511 yr You have already put a fair numbers of hours on that drive so that it should be past most of the period for an infant morality failure. I would run another full preclear cycle and check the results. If the attributes associated with ID# 5, 196, 197 and 198 are all zero, it should be good to use. If any of them are non-zero, report them back here and ask for recommendations.
February 3, 201511 yr Personally, I'd just go ahead and use the drives NOW. You can run preclear -t to see if the drive(s) already have the pre-clear signature (depending on just where things stopped, they may already be in a cleared state and ready to use) If they don't have the correct signature, just run preclear -n, which will just clear them and write the signature, skipping the pre- and post- reads. You've already exercised them enough that if there were any infant mortality issues you'd have almost certainly seem them by now.
February 3, 201511 yr Author Thanks Guys for the sanity check, Given my high confidence in HGST, that I now have about 180 hours of preclear activity if no reports), and the STRONG urgent need for a backup of my data I will add this to my array (forcing a pre-clear signature via 'preclear -n' if necessary). If this disk fails, at least the data is protected by parity. This will give me time to obtain and preclear a backup 4 TB HGST disk.
February 3, 201511 yr That's what I'd do. Note that a preclear -n will still take a fair number of hours (it zeroes the disk and then writes the signature) ... but it's FAR quicker than a full pre-clear cycle. The preclear -t will just take a second or so to check whether or not the drive already has the signature [which is actually pretty likely at this point].
Archived
This topic is now archived and is closed to further replies.