RAID controller isn't rebuilding array

delerious

Weaksauce
Joined
Jun 14, 2008
Messages
87
I have a RAID-1 array on a desktop computer. The RAID controller is a 3ware 7006-2. Recently the controller has been reporting ECC errors from one of the drives. It was able to correct all of those errors, but I figured I better replace the drive before it crashes.

I have successfully rebuilt the array with replacement drives before, but for some reason the RAID controller will not rebuild the array now. It keeps saying that the replacement drive is "Not in Service: Rebuilding" and it remains stuck at 0% for several hours. (Previously when I rebuilt the array it only took an hour or so.) I have tried using 3 different replacement drives and none work (all 3 stay stuck at 0%). I have also removed the good drive from the array and put in the bad drive (the one that had all the ECC errors) to see if I could rebuild the array using the bad drive, but that didn't work either.

Any ideas why my RAID controller is refusing to rebuild the array now? Could the controller have gone bad, even though all the other functionality seems to be working? I can boot into the OS fine, it's just the rebuilding that won't work.

One weird thing is this - the rebuilding progress stays stuck at 0%, unless I reboot the computer. Then the progress will say 1%, and it remains stuck at 1% until I reboot again. Then it says 2% and stays stuck at 2%... Maybe I should try rebooting my computer 100 times to see if the array is completely rebuilt after that.

EDIT: I know that this isn't a problem with the replacement drive being slightly smaller than the other drive in the array. Last year when I rebuilt the array, I had tried using a slightly smaller replacement drive and the controller gave an error about the drive being smaller and it wouldn't even let me start the rebuild. This time it did not display any error and let me start the rebuild, so I know the problem doesn't have to do with drive size.
 
Last edited:
Sounds like the card's port is bad, or that you bent some pins on the port when you were plugging in the new one.

An ecc error that is showing up on the drive like in your situation can can be caused by bad cables or a bad port.
 
I just replaced the cables, so it looks like a port is bad, but this is weird. Earlier today I had the good hard drive plugged into port 1 and the OS booted up fine. Then I took out the drive and just now I put it back into port 1 and the computer wouldn't boot up anymore (it said "Disk boot failure, insert system disk and press enter"). So I moved the drive to port 0 and now the OS boots up fine. But port 0 is the one I plugged the replacement drive into and it wouldn't rebuild the array.

I just ordered another 3ware 7006-2 off ebay and hopefully that will fix the problem.
 
So the new RAID controller is able to rebuild the array.

But there is an issue with Windows detecting it as new hardware even though the driver is already installed. When I booted up the computer after putting in the new RAID controller, the "Found New Hardware" dialog box popped up with the name of the RAID controller, and I went through the wizard and installed the driver (even though it was already installed). After restarting, the Device Manager showed 2 entries under "SCSI and RAID Controllers".

One of the entries is for "AMCC 3ware 7000/8000 Series ATA RAID Controller" and in the properties it says "This device is working properly". The other entry is also for "AMCC 3ware 7000/8000 Series ATA RAID Controller" except it has an exclamation point in a yellow circle on the icon, and in the properties it says "This device cannot start (Code 10)".

I deleted the entry with the exclamation point and restarted, and the "Found New Hardware" dialog box appeared again. Any idea how to remove that extra entry in the Device Manager without having Windows think that new hardware is being installed on startup?
 
Back
Top