WD Red - problem or not?

[porina]

To recap I recently had a problem on my unraid box (HP Microserver gen8) with 4x WD Red drives in it. One drive fell out of the array and unraid logs showed write errors. SMART test at the time passed. I've taken out the drive and put in a non-NAS drive and the array has rebuilt. Question is, what about the suspect drive?

 

I put it in another system and it passed Data Lifeguard Diagnostics SMART. Using an alternate SMART tool, I saw "seek error rate" was non-zero in raw value which struck me as unusual. I did both a short and long read test on it, both of which passed. Since the array had been rebuilt I didn't need the data on it so I did a full write/wipe also, which completed without error. If I look at the SMART data now, the seek error rate raw value is back to zero.

 

I'm hesitant to put it back in the raid array now, since a rebuild takes about a day. Diagnostics say there isn't a problem. People in another thread have suggest it could be a cabling problem although it was working fine before, and the new drive also seems fine. Any thoughts? @Captain_WD

 

[Master Disaster]

Have you done a block scan on the drive to check for bad sectors?

[porina]

1 hour ago, Master Disaster said:

Have you done a block scan on the drive to check for bad sectors?

Yes, as stated in OP I did a full read scan and late a full write also. Haven't done another read after the write yet. Problem seems to be seek related.

[Captain_WD]

20 hours ago, porina said:

~snip~

Hi there

 

Since the drive passes both the quick and the extended tests of WD DLG it should be considered safe to use. Could you post the raw values of the S.M.A.R.T. status? 

The WD Red drive may have gotten read error rates for a number of reasons including using it in a RAID array, using it in a NAS, making use of its energy-saving features, etc. Having only this attribute isn't really meaningful for the overall health of the drive. 

 

Captain_WD. 

[porina]

12 minutes ago, Captain_WD said:

Since the drive passes both the quick and the extended tests of WD DLG it should be considered safe to use. Could you post the raw values of the S.M.A.R.T. status? 

The WD Red drive may have gotten read error rates for a number of reasons including using it in a RAID array, using it in a NAS, making use of its energy-saving features, etc. Having only this attribute isn't really meaningful for the overall health of the drive. 

I would hope using a Red in a NAS is not surprising, since that is its reason for existing right? 

 

Anyway, following is the SMART status as of a moment ago. My concern is that even if it appears good now, without understanding why it had a problem earlier I don't feel confident to move it back into its former role and it is now just sitting as a general purpose disk in a gaming machine, replacing an ancient Green.

 

 

 

[Captain_WD]

14 minutes ago, porina said:

~snip~

WD Red drives are designed specifically for such type of usage so you should be completely fine.

 

The S.M.A.R.T. status of the drive looks perfectly fine. It shouldn't be a problem using it back in your other system.  

 

Captain_WD. 

[porina]

30 minutes ago, Captain_WD said:

The S.M.A.R.T. status of the drive looks perfectly fine. It shouldn't be a problem using it back in your other system.  

Without knowing why the past hiccup happened, here is a concern it can happen again. This could be one of those one-off things and never happens again, but I don't know that...

 

As said, for now it is put to other uses and I'll keep monitoring. I'm debating expanding the overall storage system to include double redundancy which can give more breathing room.

[Captain_WD]

1 hour ago, porina said:

~snip~

That sounds like a good plan. Check out the WD Red drive's S.M.A.R.T. status every once in a while and see if any of the worrying attributes move up or down. 

Expanding your storage for better redundancy is a nice idea. That's when you could probably use the WD Red drive again. 

 

Let me know if you have any questions! 

 

Captain_WD.