Hi all, about two weeks ago I built myself a new Server running unRAID v5.0-rc16c.  I'm having an issue where every 24 hours unRAID doesn't recognize one or two of my hard disks. What I notice is that on the main page the hard drive in question hasn't spun down and there are unusually number of high read/writes on this particular disk (but no errors).  This is usually the first tell-tale sign that unRAID is no longer able to communicate with this disk.

 

I then try to browse to this disk using a client PC and sure enough it times out.  So I then stop the array from the main page then it indicates that this particular disk cannot be found.

 

Next I reboot the server from the webUI and typically the server doesn't come back online on it's own, I even start a contiguous ping from a Windows PC but the server never replies. At this point I know that the server is hung so I have no choice but to go down to the basement and hardboot it. I run my server headless so I may have to attach a monitor to it so that I can see any console messages that appear.  I'm not certain at this point if the server is getting hung after the reboot process was initiated from WebUI or if it's getting stuck on POST.

 

The server always comes back online after a hardboot and everything works fine for approx another 24 hours until the same symptoms happen again.

 

At this point I'm not sure if I have hardware issues on my hand or if it has anything to do with the unRAID OS?

 

Has anyone experienced anything like this?

 

My server consists of: ASRock H77 Pro4/MVP MB, Celeron G540 CPU, 4GB RAM, 3 x Seagate Barracuda 7200.14 ST3000DM001 3TB 7200 RPM Drives (1 parity, 2 drives), CORSAIR CX600M PS.

Hmm, I'm not quiet sure. Do you have a syslog?

 

 

The last days I tried several times to do a parity check. This check lasts for a day or so and after this I do see errors on the parity disk and the parity check was stopped. I then stop the array and the disk is gone. It is no longer available from the dropdown list any longer.

 

If I hard reset the array the disk is back again but BLUE (new disk found).

 

 

My first thought was that I do have a problem with the disk. smartctl reports an "unknown state" of the device when the error happens. But SMART selftest and SeaTools report a good disk. In fact the disk is on its way with an RMA, just to be sure.

 

 

Now after reading your post I'm not sure if this is a disk error. My thought was the errors would lead to the missing disk. But perhaps it is the other way round: The disk is gone and the errors pop up. As I said, I'm not quiet sure.

 

I do have a syslog. This syslog was taken when this happened the first time. It was a copy right after starting the array. I didn't collect syslogs from the parity checks because I thought it would be the same error.

 

Perhaps somebody with syslog knowledge knows what happened. My arrays are 5 years old. They usually don't do such things ;-)

 

sdc_parity_syslog.txt

So after I hard booted the server the data drive in question that was not being recognized showed a red status indicator. I had to remove and add the drive to the array again and now it's rebuilding it.

 

SMART test passed successfully on the drive, I'm still not certain if my problems are hardware related or software!!!  I'm going to check the server again in 24 hours and see if these issues resurface.  If they do I will capture the syslog and post it here to see if anyone can see any issues.