I returned home from work today to find that there must have been a power outage exceeding my UPS capacity and the unRAID server was powered down.  Powered it back up, and under 5.0b13, I have it set so I must start the array manually.  All drives were green balled, and it said 'parity is valid' in the corner, so I started the array.

 

A couple minutes later I tried to access one of my drive shares in Windows and it still wasn't available.  "Odd", I thought, and flipped back to my unRAID window and refreshed the screen.  UnRAID was doing a parity check and I could have sworn it said it said it was checking '2TB' which didn't make sense as my array is much larger than that.  The box that says something like "automatically fix errors" was also checked by default, and I noticed that there were a bunch of writes happening on several disks. 

 

Concerned, I stopped the array, and started it up again in maintenance mode, and took the attached syslog so I could ask for help.  I've never really had any problems with unRAID and wanted to check here to be certain nothing is amiss.

 

Question #1: if I did see '2TB' or something like that during the parity check, is that normal?  I can't remember if it normally has the total size of the array listed there.

Question #2: why were there writes to the disks upon boot/parity check?  I've posted the activity below.

Question #3: in maintenance mode, am I safe to run Parity check now with the 'Correct' box unchecked?

Question #4: alternatively, what should I do next?

 

Thanks!

 

Device………Temp.…Size…Reads…Writes……Errors

parity….……28°C……3TB……27671…3744…………0

disk1…………22°C……2TB……24188…… 34…………0

disk2…………22°C……2TB……24199………38…………0

disk3…………24°C……3TB……25373……831…………0

disk4…………24°C……3TB……32013……669…………0

disk5…………23°C……3TB……31600……827…………0

disk6…………24°C……3TB……32365…1325…………0

disk7…………21°C……2TB……26947………38…………0

disk8…………21°C……2TB……26041………39…………0

disk9…………21°C……2TB……15889………38…………0

disk10………19°C……2TB……26720………35…………0

disk11………19°C……2TB……26802………35…………0

disk12………24°C……2TB……26490………37…………0

flash …………………  2.02GB…………………27…………0

syslog.txt

no expert but have had writes after unclean powerdown. it can take some time (15-20mins) for the shares to become available and all drives mounted.

this is the journal files replaying. which would account for writes, if that is what was happening.

 

will leave it to the experts to confirm

 

Journal transactions are being replayed:

 

Jan 17 17:50:32 Tower kernel: REISERFS (device md10): replayed 3 transactions in 1 seconds
Jan 17 17:50:33 Tower kernel: REISERFS (device md8): replayed 3 transactions in 2 seconds
Jan 17 17:50:33 Tower kernel: REISERFS (device md10): Using r5 hash to sort names
Jan 17 17:50:33 Tower kernel: REISERFS (device md2): replayed 3 transactions in 2 seconds
Jan 17 17:50:33 Tower kernel: REISERFS (device md1): replayed 3 transactions in 2 seconds
Jan 17 17:50:33 Tower kernel: REISERFS (device md12): replayed 3 transactions in 2 seconds
Jan 17 17:50:33 Tower kernel: REISERFS (device md11): replayed 3 transactions in 2 seconds
Jan 17 17:50:33 Tower kernel: REISERFS (device md7): replayed 3 transactions in 2 seconds
Jan 17 17:50:33 Tower kernel: REISERFS (device md9): replayed 3 transactions in 2 seconds

 

A few writes during files access for reading is normal in any case.

 

You should have seen 3T which is the size of parity. Run the parity check with NO correct. There should be no errors.

Thanks for the feedback - I'll kick off the parity with no correct now and let it run overnight.  (I see that you are correct about the 3TB as well.)

 

 

All appears well following the parity check - zero errors.  I stopped and restarted the array in normal mode, and data is as before the power outage.

 

Thanks for the support here; I learned a couple new things and unRAID did it's job perfectly.