-8

OK so I knew I was taking a big risk doing this but I decided to upgrade the HDD firmware on drives in my RAID5 array. And of course the controller is now complaining that one of the HDD sizes is invalid and I can't boot. I'm hoping I can just rollback the firmware version but I see no way to do this nor do I see any downloads for it online. My server is an HP ProLiant ML110 G7 with a HP Smart Array P212, running Linux (openSUSE 13.2).

HDD details:

Model: ST2000DM001
PN: 1CH164
Firmware: CC29 (was CC43)

Any suggestions, besides rebuilding my server?

Desperatuss0ccus
  • 252
  • 1
  • 4
  • 9
Mike Lowery
  • 105
  • 6

1 Answers1

4

You haven't provided good information about your hardware situation, so I'm going to make some assumptions...

  • In your comments, you mention an HP Smart Array P212 controller. Let's assume you're using an HP ProLiant server of some sort.
  • If using an HP server, you'd typically use HP-branded disks. If you're not using HP disks, there's very little reason to mess with individual drive firmware outside of the HP ecosystem.
  • What problem were you trying to solve by upgrading the physical disk firmware? Was there a preexisting health condition that could be the real problem here?
  • Do you have access to any of the HP diagnostic utilities?
  • If this is an HP server, a view of the HP IML log would be helpful. If this is just an HP controller in a non-HP server, can you post the actual Smart Array P212 POST messages that appear at boot?
  • The ideal situation would be view of the HP Smart Storage Administrator utility (hpssa or hpssacli in Linux) to see what your array thinks is happening.

Oh, and no, you probably can't downgrade the disk firmware.

But try booting with the bootable HP Array Configuration Utility Offline CD to check the status of the array. This is the right place to start since you can't get into your operating system.

ewwhite
  • 194,921
  • 91
  • 434
  • 799
  • I didn't provide all that info because I don't see it as being relevant. The question is whether or not I can rollback the firmware version. It doesn't matter what OS I'm running, why I did the update, or what server the controller is installed in. I have no diagnostic utilities available other than the controller's GUI which is very limited. I can no longer boot the server and even if I could, it's openSUSE not SUSE Enterprise so tools like hpssa and hpssacli don't work because HP doesn't support openSUSE. – Mike Lowery Mar 13 '16 at 06:05
  • 5
    The controller's error message at boot is VERY relevant. The server model was relevant because it would tell us whether you had additional logs available to you. Also the bootable HP array utility CD can at least get you to a point where you can at least see the RAID array's status. – ewwhite Mar 13 '16 at 06:08
  • "The disk drive is insufficient size and should be replaced." Pretty much what I said in my OP. – Mike Lowery Mar 13 '16 at 06:18
  • Is the array in degraded mode? Does the error message only specify one disk or multiple disks? How many drives are in the array? – ewwhite Mar 13 '16 at 06:20
  • Is this really leading to a solution? Three drives, the RAID is in failed state (hence the not able to boot statement), the error is only with one disk. And before you say to just replace it, the RAID was in the processes of rebuilding itself due to another drive having been replaced, so the data on this particular drive is critical to restoration. – Mike Lowery Mar 13 '16 at 06:23
  • 1
    I don't know the sequence of operations here, but if you ran firmware upgrades on an unhealthy disk array, it's possible that your rebuild can't complete. But it sounds like you don't want help diagnosing the issue... So maybe [data recovery is the best option](https://www.attingo.com/datarescue/server-raid-nas/server/hp-proliant.html). – ewwhite Mar 13 '16 at 06:29
  • 4
    @DiskCrasher It sounds like you're totally screwed. Dual disk failures in an R5 and data is toast. reach for your backups. – user9517 Mar 13 '16 at 06:37
  • 1
    Yes, the sequence of events boil down to: 1) failed drive that was replaced, 2) RAID in process of rebuilding, 3) server powered down and drives removed, 4) drive firmware upgraded using a different computer, 5) drives re-installed in server, 6) server (RAID controller) now doesn't like one of the drives that was working fine previously because it now shows as having slightly less capacity. I realize it's a snowball's chance in hell but if I can possibly revert to the older firmware again it just might work. – Mike Lowery Mar 13 '16 at 06:52
  • 2
    @DiskCrasher Contact [Seagate support](http://www.seagate.com/support/internal-hard-drives/desktop-hard-drives/desktop-hdd/) to see if you can revert the drive firmware. It appears to be possible. Why did you upgrade the firmware before letting the disks rebuild? – ewwhite Mar 13 '16 at 07:21
  • 1
    @ewwhite I suppose because I'm impatient and thought the risk to be low enough (not the first time I've been burned messing with array configurations). Wasn't expecting the HDD to show up as less capacity afterwards. Odder still, two drives are exactly the same model/firmware and only one changed size (according the controller) after the firmware update. The offline array utility CD claims that the data on this drive is lost, so even if I rollback the firmware I'm probably screwed. Contacted Seagate anyway. And yes I do have a (partial) backup thanks to CrashPlan, but OS reinstallation is PIA. – Mike Lowery Mar 13 '16 at 07:39
  • 7
    You have demonstrated troubling disregard for the salient concerns involved in successfully operationalizing a storage array. This is one of the simplest configurations possible and you still destroyed data. Furthermore you are profoundly arrogant in asking for help from @ewwhite who is possibly the best resource to assist in this problem. Learning from others and staying humble about what you don't know is a crucial part of your job. I take joy in knowing your data is ruined. – Joel E Salas Mar 14 '16 at 00:11
  • 5
    "I'm impatient and thought the risk to be low enough (not the first time I've been burned messing with array configurations)" - can you put a screenshot of this comment on your resume as a warning? – Basil Mar 14 '16 at 01:05
  • Great group of people. Very helpful and cheery. Makes my heart warm, as it should yours as well. Keep doing what you're doing! – Mike Lowery Mar 18 '16 at 05:04
  • (Oh, and a quick update: My RAID is now rebuilt and the vast majority of data restored. I guess you'll have to look for another source of joy there, Joel. Maybe some ISIS videos?) – Mike Lowery Mar 18 '16 at 05:11