Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

I'm trying to put together a drive configuration. I've got a server with 2 x 146GB SAS drives, and am adding several 300 GB drives.

Due to limitations on the number of drives I can put in this thing, I'd like to use only one bay for a global hot spare. I'm assuming that I can use a 300GB drive for this purpose, and it'll be able to protect all of the arrays whether they're built with 300 or 146 GB drives.

If I use a single 300GB hot spare, and a 146GB drive fails...I'm going to get a 146GB replacement from the manufacturer which isn't going to work as a global spare, so the 146GB array will have to be rebuilt a second time onto the replacement 146GB drive so that the 300GB drive can be returned to its duty as a global spare.

It would also be possible to run two hot spares, one 146 and one 300, but that would put a bit of a wrinkle in my plans because I'm at the limit on the number of drives I can put in this server.

I could also just keep a 146GB cold spare and have the 300GB hot spare only protect the arrays built with 300GB drives. The chances of failure while I'm out of town are small, and perhaps I could walk someone through the swap over the phone (yikes).

From those who have more experience trying to budget resources for hot spares, can you provide me with any insight to help me choose the correct path?

share|improve this question
add comment

1 Answer

up vote 2 down vote accepted

As to whether a dissimilar sized disk can be used as a global hotspare, it probably depends upon the RAID controller is the specific answer. I know that certain IBM Disk Storage systems will do this with no trouble, and I think the Dell PERC controllers will also do this.

On the IBM storage array I'm familiar with, when the failed disk is replaced, it automatically gets rebuilt, and when the rebuild has finished the global hotspare becomes spare again.

I don't think the Dell PERCs do this automatically, but you can do it manually through OpenManage.

share|improve this answer
    
This will be an HP P410 which I believe does the automatic rebuild you mention. I've used a couple different Adaptec controllers and an external Promise vTrak, none of which do this by default (although they can). It seems a bit silly for this to occur...except when you've got disks of different sizes... maybe this is the indirect answer to my question. –  Boden Jan 24 '10 at 4:33
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.