RAID 5 Configuration with 1 hot spare - 2 drives flashing amber - what to do?

[Gmonster]

Hi Guys,

I have a question before changing out failed drives.
I have a customer with a HP ProLiant ML370 G5
Windows 2003 Small Business Server SP 2
and Smart Array P400 Controller.

There are 4 x 146GB SAS drives in the server.  1 is used as a hot spare.
There are 2 drives in the raid 5 configuration that are flashing amber.
In the ACU there are no exclamation marks and everything seems healty.
I understand that the hot spare jumps in on a failure and holds the info till the drive is
replaced.  But there are two failed disk - does the hot spare hold the info for the 2 failed drives?
I need to replaced the failed disks but am afraid to pull them out since there are only 3 in the raid
and 2 are failing.
The server seems fine and the customer is working away but I need out move on this asap.

Can someone help me with the procedure or is it too risky - should I backup and replace server at
this stage?

Thank you in advance

[Torsten.]

Amber flshing once per second means prefailure warning - the disk still works, but has already errors and will fail soon. Replace it ASAP.

Use ACU to get the current status of the disks.

[Gmonster]

TorstenThank you - but there are two flashing amber - should I replace the two at once or one at a time?

[Torsten.]

One at a time, if you need your data ... ;)

[waaronb]

FYI, a hot spare will jump into the array if one of the drives in that array fails. Until then, it's essentially powered down.

Once a drive in an array fails, that hot spare spins up and the redundant info is written to the spare and it becomes an active member of the array.

I don't know if a hot spare jumps in if there's only a "pre failure" warning... I don't think it does.

Your best bet is to look at the array status itself and see if it's running in recovery mode, whether any drives have actively failed, and what the status of the hot spare is.

If ACU says it's all good and it's just some pre-failure warnings like Torsten mentioned, then replace one drive at a time and check in the ACU to make sure it's finished recovering the array before doing the next one.

If you pull one of the active drives, you *should* see that hot spare spring into action for a brief moment... until you put the new drive in, the P400 will just see a failed drive and use the hot spare for it's intended purpose. But once you replace the failed drive, recovery to the hot spare should stop and it'll recover to the new drive instead.

This is probably a good time to look at your ACU messages just to make sure the hot spare started to work like it should... it's always good to test these recovery plans as long as you're doing something like this anyway.

[Gmonster]

Thanks Waaronb - that is exactly what i needed to know.  I have 2 disks arriving today so hopefully it will all go well.  Thank you for your info.