RAID 1 System Disk replacement - Quick 10 points!

[SAKET_5]

Hi Michael,

I guess I am following all the guidelines that you have mentioned and the ones I am aware of. EXCEPT, there are other disks connected to the same array controller, part of different RAID sets.

Interestingly, I have used the same procedure to achieve my goal on a SAN Management appliance (Proliant DL380 with Integrated Smart Array Controller, BUT with no other drives connected to the controller).So, I am quite keen to rule out what you suggest, i.e. if there are other drives connected to the same array controller, whether it works or not!

I will give this a go tomorrow and will let the forum know.

Ciao!

[Colin_29]

The other sets on the controller will not effect what you are trying to do.

As I said, the way I do it is to shut down the systems before changing disks and run with only half of the mirrored pair at each step until I get a solution that I am happy with. Then I let the mirror rebuild.

What I said is exactly what Michael is doing its just that he worded it slightly differently.

[SAKET_5]

Hey Colin,

Sorry, I somehow forgot to assign points to your post. Okay, here is what I do:

1. Shut down the host.
2. Take one drive (out of the two drives RAID 1 set [system volume]).
3. Keep the drive taken in the above step away.
4. Boot the system with just one drive. Boots up fine.
5. Shut down the host.
6. Insert the safe disk from step2 in its original slot and pull out the other drive.
7. Reboot.
8. POST Message:
Press "F1" - all logical drive(s) will remain disabled
Press "F2" - to fail drives that are not responding - Interim recovery mode will be enabled if configured for fault tolerance.
9. I pressed "F2". I instantly got a message indicating the drive array has failed and windows would not boot. It starts attempting a boot off the NIC, assigning DHCP stuff.

What am I missing?

Cheers!

[Colin_29]

Saket,

Not sure what's going wrong here. I'm sure that is the way that I have done it in the past, I can only think that HP have changed something in the firmware and the Array controller is now recognising that a drive with a later timestamp has be used.

I will try to lab this if I get a chance later.

Sorry it didn't work for you.

Colin

[SAKET_5]

errr...not sure about the later timestamp stuff as the following is true too in the test I performed:
When I use the old two drives (that were part of the original RAID 1 set) in their original slots, the smart array detects the attachment of a *logical drive* and boots Windows fine. Thats how I got the server back.

Hope, this helps or perhaps make it more fun!

[Vijay_44]

Saket, what you are doing will only work if have one array. If you have more that one array you are going to have a problem. All array configurations are saved on the HDD themselves and are on all drives that are part of all arrays on that controller.

Now if you have two arrays and are doing this exercise, here whatâ s going to happen â

When you pull one drive out in the RAID 1 (say ID 0) â and press F2 to go to interim recovery you have up dated the array configuration on all drives (including the drives on the other array) indicating that id 0 has failed on the RAID 1. Now you pull out the surviving mirror in the RAID 1 (ID 1) and boot the server â now when you hit f2 to go to interim recovery you are updating the array configuration to indicated that ID 1 has failed â so now the array configuration has the information that ID1 and ID0 of the RAID 1 has failed and fa

[Vijay_44]

Sacket, what you are doing will only work if have one array. If you have more that one array you are going to have a problem. All array configurations are saved on the HDD them selves and are on all drives that are part of an array.

Now if you have two arrays and are doing this exercise, here whatâ s going to happen â

When you pull one drive out in the RAID 1 (say ID 0) â and press F2 to go to interim recovery you have up dated the array configuration on all drives (including the drives on the other array) indicating that id 0 has failed on the RAID 1. Now you pull out the surviving mirror in the RAID 1 (ID 1) and boot the server â now when you hit f2 to go to interim recovery you are updating the array configuration to indicated that ID 1 has failed â so now the array configuration has the information that ID1 and ID0 of the RAID 1 has failed and fa

[Vijay_44]

.............and fails the array

[Colin_29]

Well done Vijay, that explains it. I wasn't aware that drives in one RAID set held information about other RAID sets on the same controller.

This being the case, it is not possible for Saket to do what his is trying to for the reasons that your explain.

Case closed I think.

[Vincon Daniele]

Hi...


I have the same problem, resolved
using scripting utility ..
Save config of smart configurationon on floppy disk.
Whith ACR utility, you can read & save & modify & replace the ARRAY CONFIGURATION for all your purpose.

By