Hi everyone,

 

Noticed that the server shares started becoming unresponsive the other night, so jumped on the Web GUI, saw that a data drive had started to report errors. Because it was late at night, I stopped the array, and shut the server down through the GUI with no issues.

 

The following day, I booted the server again and one of the data drives had the beep of death after spinning up a little while. It was actually locking up the BIOS at startup, so I replaced it with a new drive and started the system again.

 

Now when I start up the server, the Parity Drive is showing a blue dot ! It's the same drive assignment and everything ... and the parity was valid last week when I last checked the system. The new drive is appearing in the drop down list for assignment as the replacement data drive.

 

My only theory is that the two drives were on the same output cable run from the power supply, so with the data drive failing as it did, it may have generated some power issues for the parity drive (?) 

 

Any ideas how I should approach this problem ?

 

Cheers,

 

Steve.

Post a screenshot

Sceenshot attached ....

It's been awhile since I've seen v5, but are you sure that's an error?  What happens when you assign the new drive as Disk 4?  What is the description (next to the Start button) of what happens if you click Start?  Can you give us a screen shot of that?

After you get this all fixed up you should consider moving to at t he least 5.0 stable....

I thought the same ... but the system still reports too many wrong/missing disks. ( see attached )

 

Now that I'm playing with the system, I'll upgrade to V6 after I get this issue sorted out !

 

Thanks.

 

Do you have a recent backup of your flash drive, or at least a backup of the config folder?  We're looking for a recent super.dat with that exact configuration of drives (including the missing one, not the new one).

Nothing recent I'm afraid.

 

I've attached the current super.dat and the previous backup I made. I've upgraded the parity disk and replaced another dead disk since this backup.

 

Cheers.

super.dat.zip

super.backup.dat.zip

Try the super.dat that I've attached.  It's the same as yours with one byte patched, I think, to indicate the parity drive is fine.

super.zip

Thanks Rob !

 

Unfortunately, it looks like it invalidated the configuration .... ( see attached )

 

Maybe a checksum (?) needs modifying as well.

 

My understanding of the parity drive is quite minimal, and being the technology pessimist I am:

 

Say your modification had worked, I could could start the array, "and" if the parity drive was for some reason screwed, could that corrupt the existing data on the good data drives ?

 

Thinking I should grab the good drives and snapshot a backup before attempting to start the array again ...

 

Cheers.

 

...Say your modification had worked, I could could start the array, "and" if the parity drive was for some reason screwed, could that corrupt the existing data on the good data drives ?

Parity data would only be used, along with the other drives, to rebuild the bad disk. Each disk is an independent filesystem and can be read and written independently. The only way parity gets involved is: parity gets updated when any data drive is written, and parity will be read with the other drives to calculate the data for a rebuilt disk. Reading the data from a good disk doesn't even involve the parity drive.

 

Parity is pretty simple and explained in the wiki. If you understand it, a lot of the ways in which unRAID works becomes obvious.