Failed Hard disk RAID1 x3650M2

[doggatas]

Hi

Wasn't too sure where to post this one!

I have an esx 4.0 server installed on an x3650M2.

There are two 73gb 15k SAS drives in a mirrored RAID that contains the install of ESX. One of the disks have failed. IBM have kindly sent me a replacement disk.

Can i just pull out the dead disk and put the new one in with ESX still running guests and the RAID controller will rebuild the RAID to the new disk?

OR

Do i need to migrate the guests to the other ESX server (temporarily) while i shut it down and take the dead disk out, put the new one in, boot the server up into RAID manager, rebuild the RAID, reboot server to esx and then migrate the disks guests back??

Of course either way i have to do it ill be doing it out of office hours to minimize risk!

Cheers

David

[runclear]

The Drive should rebuild automatically... But if this is a super critical box, and your simply unsure of it - I would migrate the VM's off and then run your mainteance task....

I personally have not had any issues when replacing dead drives, but i have read horror stories of multiple drives failing on a controller

-

---

| VCP[3] | VCP[4]

[doggatas]

Cheers for the answer, the box is quite critical so i think to be on the safe side i'll migrate the guests over!

[doggatas]

Cheers for the answer, the box is quite critical so i think to be on the safe side i'll migrate the guests over!

[runclear · Accepted solution]

Np, I would say this task wouldnt be any different than with your basic OS loaded on the server... If you have the ability to vmotion the guests over.. and its your job on the line.. take the safe route

-

---

| VCP[3] | VCP[4]

[doggatas]

Just so you know everything went like a charm...

I followed my steps in the second option...

Although i did have a few issues with megaRAID freezing on me and having to reboot the server to get back into megaRAID.

Apart from that it was a breeze!

I know i should document my steps for the next time, but im far too busy!