Ok, so, i found a file that looked like the permissions were/are incorrect so i decided to run the fix permissions script (running v5-RC5) overnight.

When i check in the morning, the script had hung on this HDD (Disk 7).

I looked at the main status page and it had a red mark against it.

I restarted UnRaid and powered off/on the HDD (its in a hotswap caddy) and it wouldn't redetect it.

 

Unfortunately it appeared the Syslog had done a refresh and only went back a few hours, there were no errors there.

 

I have brought the HDD to work and ran every test in Seatools for Windows.

Everything passed.

 

Attached is a screenshot of the Smart Data.

It all looks ok to me.

 

I had a similar issue with a Seagate 1 TB earlier. I replaced it with this 2TB and all has been fine so i assumed it was the drive, however it now looks it might be in the system

 

Can you please check the smart data and verify is there is a problem with the drive?

Drive looks perfect.

 

unRAID will disable a disk ONLY if a write to it fails.    You probably have an intermittent power connection, or intermittent data connection to that disk.  unRAID will not just put a disk back into service once the connection is restored since the data on it is not valid.(remember, a write to it failed)

 

You must re-construct the contents of the disk onto itself (or a replacement) to restore the disk to service.

 

The syslog files are rotated.  The older one is named syslog.1 (or possibly syslog.2 if 2 days ago)

I restarted UnRaid and powered off/on the HDD (its in a hotswap caddy) and it wouldn't redetect it.

 

Unfortunately it appeared the Syslog had done a refresh and only went back a few hours, there were no errors there.

If you checked the syslog AFTER you restarted as it implies here, then that's why. Unraid doesn't save old syslogs by default, it only has a syslog for the current session, and it's stored in RAM. If you power down without copying the syslog to a disk, it's gone.

So I just had this same thing happen on a new drive that checks out AOK.

 

if the power and data connectors are reseated how do I get this drive back in service?

 

Your notes reads like the drive needs to be formatted/precleared in a separate machine then put back in unraid server.

 

I don't have 2TB drives in my spare drawer.

 

What is the best practice for this type of failure?

 

Thanks

So I just had this same thing happen on a new drive that checks out AOK.

 

if the power and data connectors are reseated how do I get this drive back in service?

http://lime-technology.com/wiki/index.php/FAQ#How_do_I_re-enable_a_failed_disk.3F

What is the best practice for this type of failure?

See above link.  Basically, stop the array, power down, re-seat ALL connections to the drive, power AND data, drive trays, etc.

 

Then power up and follow the procedure in the wiki.  In my older server I had a power splitter that was intermittent...Drove me batty until I found it.

 

Since it must be re-constructed from the other drives and parity,  it will take a while before you are parity protected once more.  (takes about the same time as the original parity calc)

 

Joe L.

thanks for the replies guys.

I expected as much Joe, especially since the exact this happened to another drive in the same hotswap bay but i couldn't find anything wrong with it! Might be able to re-add it to the array now

 

@jonathanm, i checked the syslog in the main UI before i restarted, but didn't check on the actual boot volume if there were others created.

 

I've been meaning to do a full shutdown and re-organisation of my server for a while now...got a 5-bay caddy and a 2 port network card to install

I've also discovered i think i have my 2 current hotswap pays wired wrong. I have all power sockets connected, i read here that you should only have 2 (ie, 2 molex or 2 sata power) so will be correcting that.

 

I put the drive back in last night and let it rebuild overnight, all looking good for the moment.