Replacing an array of bad Seagate ES2 disks

[InorganicMatter]

YES, EVERYTHING IS BACKED UP PROPERLY.
(Now that we've avoided all the "back naoi" comments...)

I have a mission-critical RAID1 array here. One of the drives in it died, and after some research I have found that both of these drives contain a bad firmware that there was a huge fiasco over a few months ago. One drive is already dead, and from what I've read, the other is a ticking time bomb.

This is a domain controller, reinstalling the OS is out of the question. The array is already running impacted on one drive. I don't want to update firmwares on a live system. We have two brand new hard disks. How can I get the array safely moved to the new disks?

I'm thinking replace the currently-bad disk with a new one, and rebuild the array from the good disk. Then pull the currently-working-but-ticking-time-bomb disk out, installing the other new disk, and rebuilding again. Seems like it would work, but be messy and time consuming.

Is there a better way to do this?

 

[BENN0]

Can you bring it down for an hour or so?
Image(clone) the system with Acronis to somewhere else. Remove both disks and create a new RAID1 with the two new disks and put the image on that.
Not sure how well that is going to play if you have more than one domain controller running though.

If it is HP and you have enough empty slots to run the two new disks in the server with the old ones you can even clone the broken array to a new array depending on the controller type. You could boot the system with a recent SmartStart disk and run the array config utility from there.

But if you are running enterprise grade hardware your suggested method should work as well and if the disks are relative small and fast it should only take a few hours.

 

[fibroptikl]

Ahh, the Barracuda ES.2 firmware SN05 debacle.

Your best bet would be to shut the machine down, clone the drive, create a new RAID1 with the other drives, and then restore.