PDA

View Full Version : replace HD of smart array 3200, RAID5, un-hotswappable, proliant 800


fshguo
September 16th 04, 11:15 PM
I have a Proliant 800 server with 3 HDs on RAID 5 under Smart Array
3200, they are non-hotswappable hard drives.

Currently one hard drive is down indicated by Smart Array Utility,
first of all I would like to identify which one is dead. So by looking
into the indicator of the hard drive, there is one steadily in green,
while the other two are blinking in green together. So I am assuming
that the one with green steady must be the bad apple. Am I correct?

Secondly, the steps: should I simply power off the server, pull the
bad one out, switch the jumper with the same ID for new HD and put it
back into the same slot as the old one, then when I power up the
server, everything should be synchronized automatically with the new
hard drive and ready to go without any further configuration?

Your adivse in clarifying these steps is appreciated.

Sapian Fshguo.

NuTCrAcKeR
September 17th 04, 04:20 AM
do you have the HP System Management Agents installed ? Well, if the "array"
is telling you that the drive failed, i imagine you meant that you see this
when the machine is POSTing.

If you dont have the hotpulg drive cage, yes .. turn the machine off. One by
one, remove the drives and look at the jumper config until you find the one
with the proper SCSI ID. Replace that drive with a drive that is at least as
large, thought ideally identical. give it the same SCSI ID, and when the
server boots, the array controller will start to rebuild your volume(s)
using the new drive.

- LC

"fshguo" > wrote in message
om...
> I have a Proliant 800 server with 3 HDs on RAID 5 under Smart Array
> 3200, they are non-hotswappable hard drives.
>
> Currently one hard drive is down indicated by Smart Array Utility,
> first of all I would like to identify which one is dead. So by looking
> into the indicator of the hard drive, there is one steadily in green,
> while the other two are blinking in green together. So I am assuming
> that the one with green steady must be the bad apple. Am I correct?
>
> Secondly, the steps: should I simply power off the server, pull the
> bad one out, switch the jumper with the same ID for new HD and put it
> back into the same slot as the old one, then when I power up the
> server, everything should be synchronized automatically with the new
> hard drive and ready to go without any further configuration?
>
> Your adivse in clarifying these steps is appreciated.
>
> Sapian Fshguo.