As an experiment I've taken the drive off line and am wiping the drive by filling it with zeros which should force the drive to reallocate the bad sector. I'll do some tests and if they pass I'll re-slither the drive and see what happens.
I have sent in 6TB WD Red Pro drives and the drives I got back looked exactly like new ones, but when I sent in 4TB Red Pro drives, the replacements came back labeled as re-manufactured. The drives all looked and worked like new.
Do test all the drives you get, new or replacement, before you put them into the pool.
Like @SweetAndLow , I also keep a couple spare drives on standby. Better safe than sorry.
Yep, test, test, test, when you get a replacement. The one bad thing is that the warranty period is a continuation of the original drive so if you have a 3 year warranty and you RMA after 1 year then the replacement has two years of warranty available.
Also make sure that you package the drive correctly. The drives I purchased from NewEgg were shipped in their own boxes from the factory in a larger box with a couple "air bags" to take up the excess space. Not only did I repack the drive back into it's ORIGINAL packaging I resealed the anti-static bag and included a copy of the smart report with the issue highlighted. The small box was then wrapped in bubble wrap and boxed for shipping. The RMA said something like an inch of "foam" but I believe that would have been for a bare drive. And INSURE the drive for what it will cost to replace it that way the carrier is on the hook if it's dropped and heavily damaged or lost or whatever else could happen while they have it. I think the 150 insurance on the one I sent USPS was an extra 3 bucks. Maybe my methods are a little overkill but I would rather be safe than sorry in getting the replacement sent.
Sadly I have now been waiting about 43 hours for the return shipment's tracking information to actually work.....
Because forcing a write to the bad sector will cause the drive to reallocate the bad sector. It will no longer be a pending sector, and the SMART test will (probably) pass, so it will look like everything's OK.
The thing I normally do is replace the 'failing' drive with a previously burned in drive and then take the drive with the bad/reallocated sector to another system where I run a burn-in test again. After the test, most of the drives I have dealt with, will have hundreds or even thousands more bad sectors and that pretty definitively decides their fate. I did have a drive that cleared the single pending sector and looked healthy after and I put it back in service, but that was one drive out of dozens that survived the retest.
I'm going to replace the failed drive with a new one that has already been burnt in. When I get the RMA replacement I'll do the suggested burn in test using the solnet-array-test script and if it passes the disk will just sit on the shelf as a spare and hope the RMA drive wont fail when I need to use it.
I thought that I would post an update. I sent the failed drive back to WD as per RMA instructions. In Australia the drive had to be sent to Vietnam at a postage cost of $29.79AUD no tracking, if I wanted tracking it would of cost just over $100 more than half the cost of the drive, not worth it. I sent it on 28.10.17 and I got it back today 5.12.17, that's 36 days. You would not want to be depending on that drive. As people have stated it is a must to have a spare burnt in drive ready to go. I am currently running the solnet array test. Hopefully it all goes well.
Wow. Yes, that is quite a while to wait. Here in the US, I usually get a drive back from warranty in a couple weeks at most. Even so, I keep spares ready so I don't have to wait at all to replace a failed drive. I had some drives in one of my servers that were still running after 5 years of service and I replaced them with new drives in an attempt to prevent any sudden, unexpected failures, and when I bought the new drives, I went up from 2TB to 4TB. Maintenance can sometimes be an opportunity to get an upgrade.