RAID 5 data recovery

[Menbere]

Hi,
I have a server configured as RAID 5 .

A0-00
A0-01
A0-02
and a 4th hard disk as Global hot spare.

The problem was caused because of A0-00 failure in which case the server was expected to imdeately make the hot spare disk part of the array and start writting/building it and then resume normal operation. Unfortunately as this rebuilding process had been going on, as per my assumption, power interupption occured and the building process stoped. The hotspare disk led flashes red and is shown as failed and not accessable drive. But taking out A0-00 and inserting the hotsapre HD into where A0-00 had been shows it as a failed but an accessable one.

So now the RAId config utility shows only A0-02 as online and the two others as failed but are dispalyed as normal hard disks that can be made online if it is wished, but making them online means start rebuilding, which according to the displayed warning measge, I may loose everything on A0-02 i.e if the two reportedly failed hard disks are empty or contain corrupted information. My initial inclination is to just make A0-01 online and rebuild it. As this was a working hard disk before the event had happened I can be sure that it still holds data. This means A0-01 and A0-02 can work as RAID5 but in degraded performance. After looking at this would be positive result continue making the hotspare HD, now A0-00, online and rebuild it then I assume the RAID 5 will hence forth may operate in optimum performance mode.
Has any one advice on how I should address this problem ?

Thanks!

[Leon Rosier]

Hi,

1. i would put only A0-01 and a0-02 in the machine and power on. Normally the system will come up with the F1, F2 question. It also should find 1 logical driv(if not, it makes no sense to continue). Press F2 to fail the missing drives. If this goes well, proceed with step 2.

2. I think drive A0-00 is defective, so i wouldn't put it in the system again.
Remove all drives from the system and mark them with their position number. Put the "hotspare" in one of the bays. Boot from smartstart and do a system erase, then remove hotspare disk.

3. Update the firmware of you smartarray controller.

4. Put back drive A0-01 and A0-02, start up system , do not boot into OS (not necessary)
Hotplug the hotspare in its original spot.
Hopefully the rebuilding will start and finish now.
Put a new drive in slot A0-00 and the configuration will be brought to original.

Leon

[Menbere]

I have actually already brought A0-01 online and left the hotspare one in slot A0-00 but left it at failed status.
Server now reported as if I have a valid one logical drive and even tried to boot but couldn't lood the os. I tried to repair it using emergency disk but didn't work. So I had to boot with NT CD and supplied its NetRaid 1M driver from diskette and managed to see the Hard drive partions. I formated previous boot partition which was about 2GB and it started to load NT setup files on to this partition. I can see the other 2 data partions as well.
I tried to load NT almost 4 times now and always get "Os loader winNT couldnot start b/c :\system32\ntoskrnl.exe is missig or corrupt" Pls reinstall copy of the file."

When I looked at the Raid utility again I found out that already Hotspare HD which was in A0-00 had been made already online. I should have taken it out in the first place.
How to go forward now? I can see the data partitions but am not able to get NT or W2k installed on the drive.
thanks!

[e4services]

What I didn't understand before and still do not, is, what happened to A00-01? You say -00 failed and the hotspare started a rebuild, but then you say you only had -02 online after you moved the -04 to the -00 position? Thing is, a RAID5 needs at least 2 of the 3 drives online to have a functional logical drive and to rebuild a 3rd drive.
Anyway, you say you can see your partitions now, so you must be booted some how. So whay not get you data partition copied off? Or install an OS on a new drive and copy the data off.

[e4services]

Reading your second thread, it would seem that NT has been confused as to boot location and where it is installed with the a0-00 being online. I assume it is not included in the RAID? try and find a boot.ini in the install directory and see what it says.
As I said, at last resort, remove the RAID drives, install another drive, RAID0, install OS, place the RAID5 drive back in, see what you have.

[Leon Rosier]

hi,

Forget my story, i thought you had a proliant server, but in fact now i see you have a netserver with netraid.

Sorry,

Leon