Hello,

 

I have been running an 11disk setup since 2009 and aside from a few drive failures all has been well. Except this time. The parity drive failed so I replaced it.

I then had my mover script copy from my photo workstation while the parity was being rebuilt and I was getting all sorts of OS errors.

 

I rebooted the server and it got stuck at ACPI Tables successfully acquired, so I turned off ACPI in the bios and it booted to Unraid.

 

I looked in the log and it showed /md6 had failed. I stopped the array, got a new disk, put in the server and ran it through preclear flawlessly with zero errors. So, I added the disk to the array, and tried to format it. No matter what it will not format and throws a ton of errors in the system log.

 

The array disks are on add-in Super Micro PCI cards so then I added the drive to the server on a different sata port and I still get the same symptom.

 

I read this here -

 

http://lime-technology.com/forum/index.php?topic=32562.0

 

 

But no resolution.

 

 

I was going to swap my motherboard - any thoughts or other suggestions?

 

 

My solution was to pull disk6, do a new config init to trust the array and let the parity rebuild. I then dropped the new disk in a

all appears fine. Weird.

 

syslog.txt

I looked in the log and it showed /md6 had failed. I stopped the array, got a new disk, put in the server and ran it through preclear flawlessly with zero errors. So, I added the disk to the array, and tried to format it.

Why did you want to erase the disk by formatting it? Did you not want the array to rebuild what was on md6 to the new disk? I'm confused as to what you were trying to do.

Hope this clarifies -

 

Unraid reported /md6 as failed. So I installed a new disk and ran preclear. When I added it to the array the disk showed as unformatted. Therefore I tried to format it to try to get UR to accept the disk.

 

I would say the disk is bad but it precleared just fine.

Hope this clarifies -

 

Unraid reported /md6 as failed. So I installed a new disk and ran preclear. When I added it to the array the disk showed as unformatted. Therefore I tried to format it to try to get UR to accept the disk.

 

I would say the disk is bad but it precleared just fine.

No that doesn't really clarify. I understand what you are saying, but the question remains. Do you not want to recover the data on the failed disk? Adding a disk to the array is not the way to do that. Instead, you have to use the new disk to replace the failed disk and let unRAID rebuild the data on the replaced disk.

 

Do you actually have a redball on disk6? I'm not sure unRAID will even let you add a disk to a compromised array, but you should be able to replace a failed disk.

 

Post a screenshot.

How were you trying to add it to the array?  By stopping the array, going to the failed disk, and assigning the new disk (via the dropdown)? 

 

A pre-cleared drive isn't formatted, it is full of zeros, and has a signature written to it that tells unRAID that it can use it as part of the array.  Assuming you're replacing a failed disk, you wouldn't want to format the drive, as the drive gets it's format during the rebuild process.

You are correct, that is what I was trying to do. When I add it to the array and start the array, I get "unformatted disks" present. I would assume I would want to format that? Otherwise the array appears to do nothing?

Please answer the questions:

 

Do you have any data on drive6 that you want to recover?

 

Do you have a redball on drive6 or any other drive?

 

Formatting a drive when you are intending to rebuild a drive is not the right path.

 

Please post a screenshot.

I updated my post with a screenshot. I don't care about disk6 at all as I was able to dropped the "failed" disk into a linux machine and pull all the data off of it. The parity can be reitialized as well.

 

What puzzels me is this in the log....

 

 

un 23 19:44:24 Tower emhttp: shcmd (81): set -o pipefail ; mkreiserfs -q /dev/md6 |& logger

Jun 23 19:44:25 Tower logger: mkreiserfs 3.6.24

Jun 23 19:44:25 Tower logger:

Jun 23 19:44:25 Tower logger:

Jun 23 19:44:25 Tower logger: The problem has occurred looks like a hardware problem. If you have

Jun 23 19:44:25 Tower logger: bad blocks, we advise you to get a new hard drive, because once you

Jun 23 19:44:25 Tower logger: get one bad block  that the disk  drive internals  cannot hide from

Jun 23 19:44:25 Tower logger: your sight,the chances of getting more are generally said to become

Jun 23 19:44:25 Tower logger: much higher  (precise statistics are unknown to us), and  this disk

Jun 23 19:44:25 Tower logger: drive is probably not expensive enough  for you to you to risk your

Jun 23 19:44:25 Tower logger: time and  data on it.  If you don't want to follow that follow that

Jun 23 19:44:25 Tower logger: advice then  if you have just a few bad blocks,  try writing to the

Jun 23 19:44:25 Tower logger: bad blocks  and see if the drive remaps  the bad blocks (that means

Jun 23 19:44:25 Tower logger: it takes a block  it has  in reserve  and allocates  it for use for

Jun 23 19:44:25 Tower logger: of that block number).  If it cannot remap the block,  use badblock

Jun 23 19:44:25 Tower logger: option (-B) with  reiserfs utils to handle this block correctly.

Jun 23 19:44:25 Tower logger:

Jun 23 19:44:25 Tower logger: bread: Cannot read the block (0): (Input/output error).

Jun 23 19:44:25 Tower logger:

Jun 23 19:44:25 Tower kernel: Buffer I/O error on device md6, logical block 0

Jun 23 19:44:25 Tower kernel: Buffer I/O error on device md6, logical block 1

Jun 23 19:44:25 Tower kernel: Buffer I/O error on device md6, logical block 2

Jun 23 19:44:25 Tower kernel: Buffer I/O error on device md6, logical block 3

Jun 23 19:44:25 Tower kernel: Buffer I/O error on device md6, logical block 4

Jun 23 19:44:25 Tower kernel: Buffer I/O error on device md6, logical block 5

Jun 23 19:44:25 Tower kernel: Buffer I/O error on device md6, logical block 6

Jun 23 19:44:25 Tower kernel: Buffer I/O error on device md6, logical block 7

Jun 23 19:44:25 Tower kernel: Buffer I/O error on device md6, logical block 8

Jun 23 19:44:25 Tower kernel: Buffer I/O error on device md6, logical block 9

Jun 23 19:44:25 Tower emhttp: _shcmd: shcmd (81): exit status: -122