Error Message

[kmc]

I have a Quantum 4100 with 4 Fireball 60Gb drives, it's been playing up the last couple of days. I recieved an email alert today with the following message. Can someone confirm if this as a dead HDD or something software related. I've had problems recently with 1 drive dropping out of RAID and would hate to think this is another faulty drive....

Kaz


02/15/2006 12:51:57 FATAL ERROR PANIC : General Protection Fault (#13)
at $002303F3
--------------------------------------------------------------------

The previous message occurred in the following context within the
system log:

02/15/2006 12:51:56 ERROR File System Check : Cannot Read: Blk -16
02/15/2006 12:51:56 ERROR File System Check : Cannot Read: Blk -16
02/15/2006 12:51:56 ERROR File System Check : Cannot Read: Blk -16
02/15/2006 12:51:56 ERROR File System Check : Cannot Read: Blk -16
02/15/2006 12:51:56 ERROR File System Check : Cannot Read: Blk -16
02/15/2006 12:51:56 ERROR File System Check : Cannot Read: Blk -16
02/15/2006 12:51:56 ERROR File System Check : Cannot Read: Blk -16

[blue68f100]

This error is hd related.
Did you replace the drive that failed earlier?
Then did you rebuild the array?

[kmc]

The original failed drive has been replaced but I have been unable to get the server to boot past 75% to allow raid rebuild. I think I'm in the unfortunate position of having 2 drives die at once...

I do have a spare 4100 kicking around, dunno if there is any way I can use it if need be???

Kaz

[blue68f100]

On a raid 5. No data is lost if only one drive goes down. If you loose 2 your sol.
Did you make sure the jumpers were set corectly on the replacement drive? Was it the same size as the failed one?

There is a special procedure for rebuilding the array. I think once the drive has been replaced. It has to be setup as a spare. Then the snap will grab it and do the array rebuild.

[kmc]

Replaced drive is exactly the same manf, size, jumpers etc. Snap see's the new drive, formats it then continues booting til it gets to 75% then crashes out. It sits at 75% not moving, if you try to ping it, it seems to have dropped off the network. Might the snap be doing something, thats why its sitting at 75%??

Kaz

[blue68f100]

If you remove the replacment drive, can you connect to the snap?

[kmc]

I still have the same problem even when I remove the repacement drive, gets to 75% then bombs out which is why I think I've lost another drive. I get an em "File System Check : Cannot Read: Blk -16" although it does not specify a drive number..

K

[blue68f100]

Have you let it set over night to see if it continues. Rebuilding arrars can be slow. It has to read the data on all other drives to generate the bit that is missing.

Make sure the cpu and all cables are seated good. Received an email from one user that the cpu was not seated properly on a PV705. Not having a 4100 can not confirm.

Don't give up yet. There still hope.

[kmc]

I left it on overnight last night, came in this morning to exactly the same thing. I've been having a look round the other threads, noticed someone had problem with 'cracked raid' which is something I'm getting really familiar with here and they've used 'config devices reinit all' to return drives to original state. Is this something I could possibly try?

I'm forever the optimist but that little glimmer of hope I have is fading....

K

[blue68f100]

If the data is not needed, there are several optons.
The reint all does formating to all drives..

I like breaking the raid back to JBOD, factory def is raid 5. Don't know if the array consist of 3 drives or 4. The reason is that you can isolate drives that are bad easier and faster.

If you were remove all of the drives and use any program to format the drive, quick format ok, no unix fs. Install back into the unit and power up. The 4100 should format and prepare all drives.

There is one user MisterBill that has a lot of experience with these units. You may drop him a PM with the details. It's been several months since we have any correspondence. He was a source when I had problems.

[kmc]

I'd like to get the data off this if I can but if need be, I will have to resort to just formatting the lot. The department has a recent back up but it's more than a week old so there's probably a fair amount of data gonna be lost. I'll keep trying, maybe try MisterBill.

Many thanks for your help, much appreciated..

K