Device disabled again

Somehow, I continue to have issues using Unraid. After around one week, one disk typically disables. I am at a loss here what is driving this and previous attempts have failed to show any improvement (e.g., changed controller card, changed PSU). Any idea what could be driving this? Diagnostic attached. Thanks in advance!

tower-diagnostics-20160228-1204.zip

Somehow, I continue to have issues using Unraid. After around one week, one disk typically disables. I am at a loss here what is driving this and previous attempts have failed to show any improvement (e.g., changed controller card, changed PSU). Any idea what could be driving this? Diagnostic attached. Thanks in advance!

There are masses of Read and Write Errors throughout the log against Disk 2: (sdn) WDC_WD60EZRX-00MVLB1_WD-WX11D741AYXK

Example:

Read Errors

Feb 23 11:16:16 Tower kernel: md: disk2 read error, sector=10602791888
Feb 23 11:16:16 Tower kernel: md: disk2 read error, sector=10602791896
Feb 23 11:16:16 Tower kernel: md: disk2 read error, sector=10602791904

Write Errors:

Feb 23 11:17:06 Tower kernel: md: disk2 write error, sector=10602791888
Feb 23 11:17:06 Tower kernel: md: disk2 write error, sector=10602791896
Feb 23 11:17:06 Tower kernel: md: disk2 write error, sector=10602791904

The SMART report for the Disk looks fine and dandy. I'd try this:

Shutdown. Check the physical connections to the drive. Power on. Post a new Diagnostic File.

Thanks a lot for your reply, hugely appreciated. What do you mean by "check physical condition"? Running some form of test or just shutting down and starting up again?

He means check the cables connecting the drives to the motherboard, make sure they are properly connected and not loose at either end.

Stupid Autocorrect. Condition = Connection.

Physical Connection. Sorry.

Strike that, I did write connection and not condition! 

Essentially I mean what ashman70 said much more eloquently than me! 

Thanks a lot. I have shut down and then unplugged and plugged in the cables again. Then restarted. Please see attached new diagnostic files. Any new ideas? Thanks a ton for looking into it for hints what keep causing this issue.

tower-diagnostics-20160228-1304.zip

Thanks a lot. I have shut down and then unplugged and plugged in the cables again. Then restarted. Please see attached new diagnostic files. Any new ideas? Thanks a ton for looking into it for hints what keep causing this issue.

Nothing relating to disk2 yet. Too soon I think. Try and write / read from disk 2. Nothing important mind you BUT of reasonable size. Then post new diagnostic zip.

Disk 2 is disabled, so I cannot read / write. Or possible to mount it and then do?

Disk 2 is disabled, so I cannot read / write. Or possible to mount it and then do?

Ah, of course it disabled because it failed a write. So you have a Red X (V6.x) right? I am not inclined to advise you to re-enable the disk until we are sure that it is / was a bad connection. Not being a prat here just trying to take the cautious approach.

I am combing through your diagnostic files right now and I don't see any other issue with disk2 so I am inclined to think bad connection / cable etc. Let's see if anyone else thinks differently.

Got it, thanks. What is Red X?

Got it, thanks. What is Red X?

I have never had a failed drive before. I know that in v5 the drive status indicator for a disabled drive was always a red ball (red balled drive). As of v6.x I thought it had been changed to a red X.

Some more reading BTW re failed disks: http://lime-technology.com/wiki/index.php/Troubleshooting

Be careful before following some of that. I still say lets see if some of the more experienced users chime in and what they say.

I notice that the disk causing problems appears to be a WD 6TB drive?  I have some WD 6TB Red drives that would periodically drop offline with no apparent cause resulting in the red 'x' appearing.    I found that if I set them to never spin down then the problem went away and the system is rock solid.  The strange thins was that it did not appear to affect all of the WD 6TB Red drives that I had - just a couple of them.

I can well imagine that this may actually be the issue. How did you solve it? Letting the disk running 24/7?

I can well imagine that this may actually be the issue. How did you solve it? Letting the disk running 24/7?

I suspect that anyway the cost of having to regularly do rebuilds because a disk had been red-balled (which thus spun up all drives) was more than the cost of keeping a few drives spinning all the time.

Got it, thanks. I was less worried about the cost, but more of the lifetime of the HD assuming it spins 24/7.

You keep all your WD disks spinning or just one? Did you try to replace it or is the "slot" the issue rather than the disk?

Got it, thanks. I was less worried about the cost, but more of the lifetime of the HD assuming it spins 24/7.

I believe that spinning the disks up/down is particularly expensive in terms of lifetime cost so keeping them spinning is not an issue.

You keep all your WD disks spinning or just one? Did you try to replace it or is the "slot" the issue rather than the disk?

I keep all of them spinning (I have a total of 7 WD 6TB disks) although I think the problem may be isolated to 3 of them.  It is a trade-off as to whether to keep them all spinning or doing the work to be sure which ones have an issue and having to do a rebuild every time something goes wrong.  As far as I could tell the issue followed the disk rather than the slot although I must admit I did only limited testing on this.  Interestingly enough I also have some WD disks of other sizes in the array and they do not seem to show the issue.

Wd red drives are sold as 24/7 drives. Unless you are concerned about power usage just leave  them spun up.

I have drives that have been running 24/7 for 3-4 years with no issue and they where just standard desktop drives.

Many believe, and I'm one of them, that having a disk on 24/7 causes less wear than frequent start/stop cycles.