HP P410 Array controller / Array Lost

[aladdin24]

Hello all, can you clear out some blanks for me.

I was performing a hardware update on 2 HP ML350 servers with VMWARE ESXi installed.

I wanted to minimalize downtime of the server (S1) to be upgraded as followed.

Shutdown both servers

Take the 2 750GB SATA disks who were in RAID1 from S1

Insert them in S2

Startup S2 and configure the servers to start there.

Now the problem is when I started up S2 he didn’t see the RAID configuration, probably after I pressed F2 instead of F1 in ORCA(my bad!)

Big panic, my logical disk with my VM’s was gone!

I tried to put them back in the same slots in S1 but he didn’t recognize these as a RAID1 logical disk anymore.

I already configured a new RAID5 with 3 other 250GB SATA disks in the bays where the  2 750 GB SATA disks were so that’s understandable.

After searching lot’s of forums and didn’t find to much about it and because it was in my companies test network I took the risk.

I created a new RAID1 array using the ACU on my 750 GB SATA disks and started up, crossed my fingers booted my servers, added the storage in VMWARE and EUREAKA the datastore was still visible.

Now my question: Was I seriously lucky and doesn’t the ACU or ORCA delete Data on the disks when you create a new Array and Logical Disk?

I want to know for sure for the next time I’m in a similar situation and would like some extra information on this.

[gregersenj]

In my Opinion, at the end of the day, you were lucky.

The Smart Array controller do support drive roaming. So your plan wasn't bad.

The RAID configuration is stored on the drives only, but on all drives, that is attached to the controller.

So no matter how many drives, and no matter how many arrays and logical drives. Information (configuration and status) is stored on all drives. It's called meta data, and is stored in the RIS area of the drives.

In Your case.

You want to move drives from one server to another.

You must ensure, that the FW of the RAID controller in S2 is at the same or more recent level than S1.

Power down both servers.

Remove disks from S1.

Install disks in S2.

Power on S2.

The smart array controller will read the configuration from the mete data.

If you have put disk from bay 1 of S1 into bay 1 of S2, and same with disk from bay 2, and if cabling internally is fully identical, it will go strait through and boot, without any intervention.

If the drives swap places, due to you swapping them or internal cabling is different, it will Prompt you with a messages, telling you that drives appear to have been moved, and ask you to press F1 to continue.

Press F1 and it will update the meta data and continue.

I can't remember having seen the F2 possibility.

Only situation I remember F2, is due to boot with a missing disk (maybe also with a bad disk). Then you can choose F1 to disable logical drive(s) of F2 to continue in interim recovery mode, F2 is defaulted within 30 secs.

If logical drives has been disabled, they can be re-enabled, through a reboot or from the ACU.

If you delete a logical drive and array, it can't be re-created (To my best knowledge).

On S1. You removed the drives, put in new drives and created a new RAID. Using the same physical positions

No problem, simply power down IS remove, the "new" drives. Insert old drives, and it will pickup the configuration from the "old" drives.

Here's some good reading about the smart array:

http://h20000.www2.hp.com/bc/docs/support/SupportManual/c00687518/c00687518.pdf

I don't know where you got it wrong, or how got it right again, but I believe you had some luck.

I have also done some, not so clever things, and have been saved by the smart array.

BR

/jag