1

I searched but found no solution to this question.

I have a HP ProLiant DL385-G5p server with RAID 5, having 4 hard drives: 3*75GB + 1*75GB spare. One of the 3 hard disks in RAID5 has failed (the indicator shows this), and the spare is up an running.

My question is: can I replace the failed drive with a 150GB hard disk, by unplugging the failed drive while the system is running, and hotplugging the bigger 150GB hard disk? Is there any risk of system failure? My guess is that the array controller should rebuild automatically, but one never knows.

Thanks!

Improve this question

3 Answers 3

Reset to default
2

Can I replace the failed drive with a 150GB hard disk, by unplugging the failed drive while the system is running, and hotplugging the bigger 150GB hard disk?

Yes, the array will begin to rebuild using the new disk.

Is there any risk of system failure?

No more than when replacing the disk with one of the same size.

Improve this answer
2
  • Ensure that the drives support hotplug. Not all HP enterprise drives do.
    – Terry Burton
    Jul 30, 2015 at 13:01
  • Very true. However, he's talking about an HP DL385 G5p, and while HP does offer an option to use LFF drives in the 38x series, I've yet to ever see one. Hence my assumption that he's using SFF drives, which are all hot-plug.
    – GregL
    Jul 30, 2015 at 13:12
1

You can do this, but I'd recommend finding a like-model disk rather than using the larger 146GB 2.5" drive.

If you have no other options (especially since 72GB disks are no longer manufactured), the 146GB disk will work and rebuild. The size of the disks and the vintage of the server indicate that this is an older G5 system (2005-2008), so at some point, make sure you upgrade the firmware on your Smart Array P400 RAID controller as well.

Improve this answer
0

This works just fine on our HP ProLiant G6 servers with P410i controllers. As I see it, it can be necessary to do the switch while the server is online, so the controller knows which disk is the active disk (i.e. in RAID1).

If the reused disk is switched while the server is offline, the controller normally might read the meta data from the disk when starting up and could get confused about which disk to rebuild. I have not tried this.

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .