I went to sleep last night with a few things going on. 

 

My 6yo came in the room and did "something" but when I got here, the server was off.  He said he pulled the usb cord going to the UPS, but not sure how that would make the server go off.  At first it wouldn't turn on either from the power button.  Not sure what was going on I unplugged everything and plugged the server directly into the wall, and it started to power up, so I immediately stopped it (holding power button) before it really booted and put it back in place.  I power cycled the UPS turned the server back on and this time it started up just fine.  Of course it wanted to run a parity check, I did stop it to do a couple other things and then restarted before I went to bed along with two preclears running.

 

When I got up this morning, I have this screen (see attached screenshot).

 

The syslog is 5000 lines most of it repeating.  I cut the majority of the last part of the syslog to get it uploaded.

 

I could not get a smart report to run on the drive.  I will power it down and check cables to make sure something isn't loose.  power back up and see if anything is different. 

 

Questions:

 

1. It says parity check complete with no errors, but I figured this occurred during the parity check, so a little confused there.

2. It is showing some writes to the failed disk and the parity disk (see screenshot).  The writes to disk 8 are probably due to me opening a file off that disk this morning.  Is there any chance the parity was corrupted having a drive fail during a parity check and still show up as valid?

3. I have a spare drive already precleared here.  Is there any reason not to just go replace the drive?  Anything else I should check first?

 

Thanks, will post back anything I find when I check cables and reboot.

 

Hardware list and unRAID version in my signature are current.

Here is the syslog.

 

noted where I deleted repeating lines.

syslog-2012-03-24_cut.txt

I powered down, checked all the cables (which seemed perfectly fine), and rebooted.

 

The disk is still disabled, but I was able to run a smart test.  Attached are the status reports both before and after running a smart test.  Also the syslog after reboot.

 

The disk is a little noisy running the smart test.  Not really loud, lots of rattle head noise, but not like it was when new.  I think I've noticed one of the hitachi drives has been much noisier lately than it was when I first bought them.  Not sure if this is the same or not.

 

2012.03.24_Smart_Report.txt

syslog-2012-03-24_after_re-boot.txt

The SMART report looks ok except for a single pending sector. Rebuild the drive: Start the array with the drive unassigned. The assign the drive and restart the array. The drive will rebuild.

The SMART report looks ok except for a single pending sector. Rebuild the drive: Start the array with the drive unassigned. The assign the drive and restart the array. The drive will rebuild.

 

Rebuild the same drive or replace it and rebuild?  Is there any way to "test" and make sure it is reliable?

 

I could install the new drive in place of this  drive, rebuild.  Then preclear this one again?  Or based on what you see is this not necessary?

The safest course is to rebuild with a pre-cleared replacement then the pre-clear the problem drive to test it's reliability.

The safest course is to rebuild with a pre-cleared replacement then the pre-clear the problem drive to test it's reliability.

 

I went ahead and swapped the drive for the new one last night and let it rebuild.  Everything up and running fine.

I will run a preclear on that drive, probably with 3 cycles next weekend. and see how it goes. 

 

Now I need to find a deal on another standby drive to keep on hand.  Good HD deals are still hard to come by.  Seen a few at $110 for a 2 TB.  I'm thinking I want to start buying 3TB drives but they are still outrageous.