Hi,

 

Should I be worried or take any action for the following?

 

Have a relatively new build that was working fine with absolutely no problems.  Running a completely stock unRAID Pro 4.5.4 that had a normal parity check 2 weeks ago (0 errors).

 

I was copying a large 7GB file over my giganetwork from my Win XP to unRAID, as I normally do without a prob, when I got a network connection error.  All of the sudden my network shares for unRAID not visible and //tower console not coming up.  The network is wired through the house.  Since the unRAID is headless I looked at the KVM screen, but didn't see the usual tower login prompt.  It wouldn't let me do anything so I pushed reset on the unRAID and it rebooted pretty normally.  After it rebooted it appeared normally back in the windows network and web console worked as well.  A few of the disks took a few extra seconds to mount, but mounted ok (as noted from refreshing the the web console screen a couple of times) then it automatically started a parity sync check.  I deleted the file that I had been copying over the windows network.  (It was reported as the whole file but I knew it wasn't).

 

The parity check just completed and with 384 errors.  None of the disks reported any errors.

 

What does this mean?  Should I do anything further?  All my disks had been precleared Samsung 2TB originally with no problems.  Is the parity now good again?  Can I assume the data on the drives are still good and the errors are from the sudden connection loss, then me deleting the file?  The system seems to be operating normally and streams my media normally as well.

 

Thanks for any input!

 

 

 

 

Hi,

 

Should I be worried or take any action for the following?

 

Have a relatively new build that was working fine with absolutely no problems.  Running a completely stock unRAID Pro 4.5.4 that had a normal parity check 2 weeks ago (0 errors).

 

I was copying a large 7GB file over my giganetwork from my Win XP to unRAID, as I normally do without a prob, when I got a network connection error.  All of the sudden my network shares for unRAID not visible and //tower console not coming up.  The network is wired through the house.  Since the unRAID is headless I looked at the KVM screen, but didn't see the usual tower login prompt.  It wouldn't let me do anything so I pushed reset on the unRAID and it rebooted pretty normally.  After it rebooted it appeared normally back in the windows network and web console worked as well.  A few of the disks took a few extra seconds to mount, but mounted ok (as noted from refreshing the the web console screen a couple of times) then it automatically started a parity sync check.  I deleted the file that I had been copying over the windows network.  (It was reported as the whole file but I knew it wasn't).

 

The parity check just completed and with 384 errors.  None of the disks reported any errors.

 

What does this mean?  Should I do anything further?

First the parity "check" is not just a "check." It is a "check and repair" so once it tells you of the problem it has already corrected parity to match what it read from the data disks.

 All my disks had been precleared Samsung 2TB originally with no problems.

The previous clearing of the drive has absolutely nothing to do with what is happening now, other then it indicated the drives were working properly at that time.

 Is the parity now good again?

Yes. It is in agreement with the data on the data disks

Can I assume the data on the drives are still good

Probably as the file system used is a journaling file-system.  When you re-started the journal entries were re-played before the disks were put back online (and why it took a few more seconds to mount).  Some users with heavy file activity prior to a crash have reported it took 5 or 6  minutes to replay the journal and before the disks would mount)

and the errors are from the sudden connection loss,

Not the connection loss... That was a symptom, not a  cause... Your server crashed.  That was the cause.  We still do not know what caused it to crash.  Even though the data disks have journals to replay their contents, it is possible for some of the housekeeping data to be out of sync on the parity disk, those are likely the "errors" you saw on the parity check upon reboot.

then me deleting the file?

You deleting a file should not cause any errors.

 The system seems to be operating normally and streams my media normally as well.

 

Thanks for any input!

That is good.  

 

I'd check the obvious...

Look for errors in the syslog.

http://lime-technology.com/wiki/index.php?title=Troubleshooting

Perform a memory test, overnight.    If it has errors, nothing else is going to work reliably.

Check your disk file-systems.   I've seen them initiate a system crash if they are corrupted.

http://lime-technology.com/wiki/index.php?title=Check_Disk_Filesystems

 

Joe L.

Thanks, very helpful.

 

smartctl reports normal and reiserfsck show no corruption.

 

I'll perform the memtest tonight.

 

Thanks for your input.

syslog.txt