DL380 Not Booting, Urgent, Help!!!!!

[Carlos_116]

Dear all:

Server is a DL380 G1 with an 4200 array controller and a Raid-5 array of 4x18GB. Running Windows Small Biz Server 2003. Server has been running perfectly fine for a long time, not errors in the array.

I was upgrading the disks. Replaced disk #3 hot (on top) and it rebuilt without a problem.

When disk #3 finished and I saw in Insight Manager that everything was green,
I pulled disk #2, the array kept on working and all of a sudden disk #1 gave me
a red cylinder light (the one in the middle). PANIC, two disks off-line!!!

I put disk #2 back in (the one I had just pulled out) and nothing happened. I then
powered down and rebooted. On bootup the array controller said that disk #1 needed to be replaced, which I did do. Based on the lights on the drives I thought that the array was rebuilding disk #1 (blinking green cylinder), but after what I thought was a short period of time for this process, all the cylinder lights and the arrow lights went green on all the drives. This is steady green, no blinking whatsoever. At the same time, there is no activity on the disk light of the server itself and I don't hear any head movement in the disks.

After waiting a while, I rebooted. The controller initialized just fine and did not
report any problem with any of the drives but, I got an error saying that could not load ntoskrnl.exe.

I then rebooted to the SmartStart CD and ran the array diagnostics. It came back clean, no problems found, none at all.

I reboot again, I now get the Windows text based interface to choose the booting O/S, the dialog you get when you press F8, but I get without touching anything. I tried it twice with reboot with last good configuration and again with normal and nothing happens, no error, cursor just stays in the upper left hand corner. The disks still have the cylinder and arrow lights steady green.

What is going on, is my array dead????

Your help will be greatly appreciated

[Oleg Koroz]

HOW you pulled drive out when System was up and Running?
Or you brought Down Server?


Sound like you corrupt system files or mbr ...

[Oleg Koroz]

see also attachment

[HGN]

Hi

Can understand that you swapped the disk 3 since it was hot spare ,but disk 2 when the server is up might have got the data corrupted.

Rgds

HGN

[kcpant]

Hi Carlos,

I'm not very clear how these four disks are configured, is the disk #4 is a hot-spare, or is it also a part of your RAID5 set? you have only mentioned in your post about disk #1,2 & 3, not bout #4...

[e4services]

They data has been lost. Though disk #3 rebuilt without error, there most have been a problem with disk #1 all along. So when you tried to swapped out disk #2 and while out, the contoller now saw some problem with disk #1 and the array and the data, it took disk #1 offline.
Well once you have 2 drives offline, the array fails. What you did after that is not going to effect that fact that the data is lost and must be restored.
This is why a good backup is needed before starting an online upgrade of the disks. Anything can happen in this long process to cause data loss.