0

Edit: I upgraded Open Manage Server Administrator and now I receive the following message when I attempt to assign the new disk as a global hotspare:

Caution

Physical Disk 0:0:0: This Physical disk is not suitable to protect all current or potential virtual disks on this controller. Possible reasons include: Insufficient physical disk space, unsupported mix of SAS and SATA type physical disks, unsupported mix of SSD and HDD type physical disks, unsupported mix of 512Bytes and 4KBytes sector size physical disks, unsupported mix of PI capable and incapable type disks and non-SED drive assigned as a GHS when encryption virtual disk is present. Continue with assigning as a global hot spare?

So it seems there is something about the disk I need to configure first. I will post the solution once I figure it out.

Original Post

I have a Dell PowerEdge T410 server with Windows Server 2008 R2. Storage is configured in a RAID-5 array. Everything has been working fine for several years, although I've had to swap out failed drives a couple of times here and there and rebuild.

Until now, rebuilding has not been an issue, and the Open Manage software makes it pretty easy to manage. With the latest drive failure, however, I am seeing something different.

I bought a new drive (same manufacturer, model number, and capacity as the other three in the array), and after popping it in, I deleted the virtual disk automatically created for it and assigned the disk as the hot spare for my array. From there, the rebuild is seemingly initiated, but progress never goes beyond 0% (it has been about 18 hours since it was initiated).

Here is what I am seeing:

Open Manage Virtual Disk(s) on Controller PERC S300

When I click the Name of the Virtual Disk (SERVER) to view the physical disks, this is what I see:

Physical Disks included in SERVER

I've checked windows logs and don't see any failures or errors related to this within the past 24 hours (when I first swapped out the disk).

Looking at the OMSA logs, it tells me when I unassign a global hotspare, but there is never an event corresponding with me assigning a global hotspare. Similarly, I see notices when I cancel a rebuild but never when it is initiated after assigning a global hot spare.

I've done some looking into similar problems, but most people receive an error or an actual failure during rebuilding, which I am not.

I have good backups and the system is still running okay, so my next thought is to just take an image, blow away the array and start from scratch. I am mostly looking for other ideas before going that route.

Community
  • 1
Gage Trader
  • 161
  • 1
  • 8

2 Answers2

0

Unassign the disk as a global hot spare and leave it unassigned. The RAID controller should automatically pick it up and rebuild. Very odd your system build a VD immediatly when you inserted the new disk. In my experience when you put a new bank drive in a Dell with a failed drive it automatically picks it up and starts rebuilding unless there was preexisting data. Might want to also try clearing the disks config completely in the PERC and then remove/reinsert it and see if a rebuild automatically occurs.

You could also be running into firmware issues. Looks like your new drive has a newer firmware than all existing drives. Might be worth upgrading them all to the same version and your PERC controller to. I've never seen firmware cause an issue like this but it couldnt hurt.

I believe Dell iffers support for the life of the server. They just wont send you parts if your warranty is expired and you end up needing them. Give Dell a call and see what they say.

FiZi
  • 84
  • 4
  • Those are all good suggestions, thanks for the reply. I did try leaving the disk unassigned, but OMSA was still showing it in "rebuilding" state at 0%, even with the disk unassigned. – Gage Trader Dec 19 '15 at 19:36
  • Dell doesn't offer lifetime support. The server comes with whatever support contract was selected at the time of purchase, typically one to three years. Once that support contract expires they will not provide support. That's why it's important to renew the support contract. – joeqwerty Dec 19 '15 at 19:37
  • Oh interesting. Ive called before with expired systems and they've offered support up to replacing parts. Maybe its because we usually buy ProSupport or maybe theyve changed something. I did call them about a Dell PE2650 2-3 yeats ago and got support on a BIOS update issue. That server was 8 years old at the time.... And is still with us today :/ – FiZi Dec 19 '15 at 19:39
0

Okay, so I'm still not entirely sure what solved this problem, but the array is now rebuilding.

The first thing I did was to remove OMSA completely and do a clean install (when I upgraded before, I went from 7.4 to 8.2 using an upgrade package). After uninstalling OMSA 7.4, I rebooted the server and went into the PERC S300 array manager that is an option on boot.

There, it was showing Disk 0 as Ready and Disk 1 as being the Spare (OMSA did not show disk 1 as being a spare).

I set Disk 0 to the Spare and unassigned Disk 1 from the Spare.

After that, I proceeded to boot and install OMSA 8.2. After installation, I went to view my virtual disk and viola, it was finally rebuilding.

Gage Trader
  • 161
  • 1
  • 8