Western Digital RED failure rate and poor quality re-certified replacements

[cyber69]

Get back from a short break this week to find my 4 bay Synology is having another issue with a 4TB red drive. So I have a offline spare which I keep on standby and replace the failed drive and it fails to rebuild saying drive 3 has bad sectors.

I also find that drive 2 has had 2 re-connections in the past month meaning that it won't fail just yet but probably at some point.

So that's three drives all with issues and they are ALL replacements which I have received from western digital for failed drives. It seems that you cannot put any trust in re-certified drives in a raid.

If you own a Synology unit and on the monthly extended test a drive appears to be stuck at 90% then it DOES have bad sectors and needs replacing.

I'm stuck now, having to RMA two drives back to WD and one drive is out of warranty. I think i'm going to put two 4TB's back in a raid 1 this time and run two drive's in a JBOD with less critical data on them. I just cannot afford to replace all 4 of them at around £480.

 

[mikedigitales]

I've not had this problem yet with the 4tb Reds, I have had a few recertified WD in the past (3tb Reds) that survived a few years before I moved them on. So you might just be unlucky? How are your average and peak drive temps?

 

[cyber69]

32deg temps average. I bought this DS916+ some years ago and all the original 4TB red's have now failed. On a DS418 I bought with 4 * 4TB reds none have given errors (yet - touch wood) The problem i'm getting is with the re-certified replacements with two being DOA also. So far out of the original 4 drive's i've had to raise 6 RMA's against failed drive's.

 

[cyber69]

Hmmm, managed to take 7TB of data off onto USB storage temporarily and took each drive out and tested on WD diags and all passed? Put them all back in and powered up and the Synology came up with 'DISK 4 FAILING!' so I guess it is detecting something more than WD diags or has a lower threshold for failure tolerance. So that drive is going for RMA and put my spare back in and it's currently rebuilding ok and not failed quickly like yesterday.

Taking the data off the raid means less to rebuild I guess. The reason I went for 4TB drives is less to rebuild and more chance of data recovery unlike 6TB+ drives. I certainly would not want 10TB drives to rebuild and that comes from 30+ years working in I.T.

 

[EsaT]

If drives seem to work elsewhere maybe that Synology has power problems?

Anyway butterfly read should be pretty good stress test for mechanisms of whole drive if you want to see if they're starting to break.
http://hddscan.com/

 

[cyber69]

Got drive #4 replaced all ok after moving 6TB of data off onto backup drives. I guess no data to rebuild from the bad sectors helped. RMA'd the faulty one back to WD and got the replacement dropped off last night by UPS and used that to replace drive #3 which was working but has over 48 bad sectors flagged. Left it running and it rebuilt ok but now got an error with drive #2 failing.

So looks like all 3 re-certified drives are faulty.

 

[RobTi]

Had this on my home built whs with drives failing and testing ok out of that machine, turned out it was the psu on its last legs,even though I tested the output on all rails, swapped it over for a spare and it’s been fine since then

 

[cyber69]

Drive #2 replaced now and raid all rebuilt ok.

Ran an extended smart test on all the drives, 3 & 4 are all clear but the new warranty drive put in to replace drive 2 has some error



The multi_zone_error_rate reads 6 which is still a pass on the extended tests.

On drive 1 which is still the original WD red 4tb this flagged up some issue also



One multi_zone_error_rate reading



And 3 raw_read_error_rate readings

Both drives say zero bad sectors and zero reconnection counts and pass the extended tests ok.

I'm copying 3TB of data back onto the raid now and will see how it goes.

I've got the final RMA drive to send off and receive back so I will have a spare again to replace either drive 1 or 2 but I think
drive 1 will be the next one to start giving errors.

I've got 4 x 4TB red's in my DS418 play NAS and they are about 2 years old and have been 100% solid (so far). I guess i'm having
issue's with the crap reconditioned WD drives they send out.

The RMA packaging they send the replacement in all the way from Poland is a basic cardboard box with a couple of cheap plastic holders for the
drive.

 

[mikedigitales]

Have you tired running a few passes of badblocks on the replacement drives before using them? It's a great way to root out any damaged in transit drives before rebuilding your array. I think the some of the SMART attributes you mentioned can go down as well as up, so might be nothing to worry about unless they fail a long test.

 

[cyber69]

Took out the latest replacement drive and tested it on WD diags but on extended test and not a quick test and after 10 min it seemed to stick reading one sector on the drive and then fell over with 'too many bad sectors'. So i've had to raise another case with WD to get this DOA drive replaced. In the meantime i'm moving all the data back off onto USB storage again.

Seems strange it passes the Synology extended tests and failed the WD extended test. That probably means that I will have to run the WD extended tests on all the drives received from now on.

 

[cyber69]

Took drive 1 out which is last of the 4 original drives left and ran WD lifeguard diags on it and it failed with 'too many bad sectors' so that means that one will need replacing also. WD won't allow me to raise another RMA on that one as i've logged 2 calls on the same model of drive in the past month so i've had to create a support case.

All my data is off the NAS and on USB storage so I don't know if to call it quits and start selling these recon drives on ebay to get rid. I have 2 drives which seem to be ok and pass all extended tests so these could be sold straight away and I will eventually end up with 3 more recon drives I can sell.

 

[cyber69]

Bit of an update, WD created a manual RMA for drive 1 and organised a replacement drive for the DOA one. I’ve learned my lesson by just trusting a quick drive test so got out an old laptop and connected the drive to it by USB then ran an extended WD lifeguard diags test which took 9hrs.

It came back all good so this replaced drive 2. Now need to send the DOA back to WD in Poland by UPS but they provided a paid waybill for this.

Just got the replacement for drive 1 on extended tests now. If that passes then that will go into the NAS and rebuild and once complete I will run a Synology extended test on both the drives.

I will finally run data scrubbing on the pool.

 

[cyber69]

That's all four drives replaced now since returning from my break to find it borked.

All pass the extended tests so copying the data back on now.

If you own a Synology unit I would recommend running the smart extended tests once a month to check on disk health properly. If a disk seems to be stuck at 90% then it WILL have bad sectors so my advice would be to power off the NAS and connect it to a PC and run WD LC diags on it using extended tests and that should fail with an error 'too many bad sectors'. Then log it as an RMA with WD to get it replaced.

If you are thinking of buying a NAS then my advice would be to buy a spare hard drive for it at time of purchase (if more than a 2 drive raid 1) so you can use this as an offline spare. When you have a drive failure you can replace it straight away and RMA the faulty drive.

And always run an extended test on any re-certified replacement drive coming back before putting it into the NAS.

 

[cyber69]

All good so far but still waiting for my last replacement drive to be delivered. Drive was sent in last Monday and received on Tuesday and RMA was updated to show faulty drive received on Wens and then nothing. So raised a case with WD and this is what they came back with:=

Please note: Due to year end physical inventory count during the week of June 24-June 30, 2019 no receiving or shipment of replacements can take place.
Thank you for your cooperation and we apologize for any inconveniences this may cause.

We expect normal shipping to resume on Monday 1st of July.

Not good really if you needed a replacement quickly even if it is on a standard RMA. I guess an advanced RMA would have shipped? Anyway it is for my offline spare so no immediate rush but nonetheless....

 

[wonko]

It's a while ago now, but I also had a bit of a nightmare with failures of WD reds and their recertified replacements, although not as bad as the OP's case.
The first drive that developed problems was replaced ok, but the second turned into a bit of a nightmare. Advance RMA wasn't available for some reason, and when I received a faulty recertified replacement, WD agreed to the next replacement to be a new drive. So when another recertified one arrived it then took weeks to get them to honour what they had said. Part of the frustration was that there was no escalation / complaint process (or if there was, they wouldn't tell me about it).
From that experience I'd prioritise support over any marginal differences in failure rate between brands/models.
Meanwhile, I've got a 6TB red that has dropped out a couple of times...

 

[Goksly]

I've got a 2 month old 4TB WD red drive with bad sectors which it's in a Synology NAS. It's got 10 bad sectors; is this just the beginning? Should I be thinking of starting a RMA or so I need to wait for it to be a lot higher?

 

[mikedigitales]

Depending on where you picked it up Goksly I'd be looking to return/replace it with the retailer I think - no need to muck about with getting a recert if it's that new? I'd definitely get it sorted straight away though, any bad sectors would make me get it sorted asap.

 

[Goksly]

Purchased from rainforest... But saying returns window closed a month ago (a month after purchase)? Cant be right for faults that surely?

 

[HungryHippos]

Purchased from rainforest... But saying returns window closed a month ago (a month after purchase)? Cant be right for faults that surely?

Go on their live chat, they mean returns window for "unwanted items". Yours is a faulty one, their live chat people should send you a new one out, you can then return the old one after usually.

I am replacing my 8 x 3TB Red drives (SHR - so only 1 disk redundancy) with (up to) 4 x 10TB Red drives.

Currently my raid set is fine but 2 of the disks have 1 bad sector each. These have had bad sectors for a while now and doesn't seem to show any signs of worsening.



I'm copying stuff off in preparation for the replacements to go in, at the moment I am toying with the idea of just putting 3 of them in in SHR, so I should get 20TB usable space. That will let me add the 4th one in should I need more space, or if one fails.

 

[cyber69]

^ With bad sector errors you might find when replacing a totally failed disk that the volume fails to rebuild because it cannot read data from the bad sector on another disk. In that case try moving data off the volume and restore from backup the corrupted files.