HP Proliant Microserver | AMD RAIDXpert RAID Rebuild

I recently developed a few faults with Windows Server 2008 R2 install running on my HP Proliant Microserver.
A number of posts and articles suggested the fault maybe the result of a faulty HDD therefore I ran a series of disk checks and applications to verify. Unable to locate any faults I opted to remove both 2TB HDDs (configured in RAID1) in order that I could dock them in another workstation and run thought some more thougher checks.
strangely as a result of removing the HDDs the Windows “BSOD” and crashing did stop, however I was still unable to locate any errors on the HDDs.

Rather than installing both HDDs back in the system, I thought it best to only reinstall one HDD in order to fault find, therefore breaking the RAID. Once happy that the fault was correct (Never found out the issue) I started RAID1 rebuild process using the following steps.

Before I physically installed the second drive I launched the AMD RAIDXpert Utility to see the RAID status, clearly showing as “Critical” due to the missing drive (Only 3 drives are listed)

Although the Microserver doesn’t have hot swap drives, I reinstalled the second HDD (without shutting down) simply running a “rescan” in “disk manager” to get it to show. Once back in the RAIDXpert tool (with a refresh) the installed HDD was listed. Clicking on the HDD displays the properties. The point to note here is the colouring around the drive. This is the HDD which is currently working, the status is “Assigned” as it is part of the “Data-R1” RAID set

This is the HDD which requires a rebuild, the status is “Available” as it is NOT part of any RAID set. (I assume the Raid becomes broken when the data on the HDD differs from the original by a lot or if it has been removed for a specific time period.)

To rebuild the arrange I selected the RAID set and entered the rebuild section, as the 2nd HDD has a status of “Available” this can be selected and the rebuild started “Start Now”

The RAID rebuild process starts… This process takes a long time, actually days!

Luckily the drive can still be accessed as normal in Windows, while the rebuild is taking place.

3 days on and the rebuild process is still only on 60% so I wouldn’t be waiting around for it, just let it do its thing!

Finally it should reach the 100% mark, hopefully giving you peach of mind that all the data is now duplicated once again!

3 thoughts on “HP Proliant Microserver | AMD RAIDXpert RAID Rebuild

  1. Nigel

    Hi Dom,

    Thanks very much for this! I also have an HP microserver, but running Win 7pro, and wonder if you can help? Mine has 2 pairs of disks in RAID1 – bay and and 2 paired as Logical Drive 1 with partitions C: and D:, and Bays 3 and 4 paired as Logical drive 2 in Logical Drive 2 called U:

    I had a failure in bay 2 so logical drive 1 went critical. I couldn’t get hold of the original WD20EARS drives, so bought 2 WD20EZRS (6Gb not 3Gb SATA transfer and WD said slightly different spin speeds as the firmware version is different. They advised to rebuild both immediately as they would fail within a month). I put the drive in and it seemed to come up with a white line around as ‘avaliable’ in RaidXpert without formatting, so i tried to rebuild as above but get multiple bad sector errors… is this because its not formated, a duff new drive, or because the drives aren’t quite identical please? Stupid question but if i need to format do I format during booting? I cant format one bay in drive manager as they are now paired as logical drive 1. Alternatively do I put in another new drive, format and then add to LD1 and rebuild? I could then get a caddy, and manually format the first new disk, swap the rebuilt EZRS from bay 2 to bay 1 and then rebuild a new formatted drive in bay 2 again?

    When the RAID is rebuilt, does that automatically build the partitions or do I have to manually create those after formatting? I have sys reserved 100MB, C: 146.48 GB NTFS, and D: 1716.37 NTFS C obviously has the OS on too….

    thanks so much for any advice you can give !

    regards,
    Nigel

  2. Dom Post author

    Hello Nigel, apologies for not replying sooner (I’ve been on Holiday).

    Usually you don’t need to format a drive before you install and rebuild the RAID array but it can be good practice to do so (This is why there is a initialize option but that wipes everything). It’s hard to say if the HDD is physically faulty but my guess would be that it is a read/write error (There’s some info here about Hard/Soft Bad Sectors (http://www.howtogeek.com/173463/bad-sectors-explained-why-hard-drives-get-bad-sectors-and-what-you-can-do-about-it)

    I’ve rebuilt a number of RAIDs using different size disks (These have even been different sizes supplied by the hardware vendors i.e. Dell & HP) so this shouldn’t be an issue, you will just loose the additional space if larger. Spindle speed (5400 or 7200) might be a different issue, I would always look at using the same RPM, however you can use different ones but it will only run at the slowest speed.

    I would be very careful with the good HDD and leave it in place. You don’t want the RAID to rebuild using the blank disk as the source otherwise you may wipe everything. As the clone is completed by the Hardware it will take effect as soon as you boot the system and NOT after you have loaded Windows (Software RAID). Best option is to remove the new HDD format in another device (if possible), reinstall and let the RAID rebuild again, hopefully the format will clean any bad sectors making it successful this time. You could possibly have bad data on the old HDD which is being replicated so something else to be aware of. If you don’t have another system to format in then best option would be to use a USB to SATA adapter.

    When the RAID is rebuild it will automatically handle all the partitions etc. you will not have to manually recreate any of these. The hardware RAID makes a complete replicate of the source HDD at block level (https://en.wikipedia.org/wiki/Block_(data_storage))

    I’m always cautious when down to 1 HDD as there is potential to lose all data so it’s important to do a backup. Usually this can be done by booting the single HDD and running a Windows backup before attempting any fixes etc.

    Hope that helps a little.
    Let me know how you get on.

    Thanks,
    Dom

  3. Nigel

    Dear Dom,
    Many thanks for your reply! However, I think it probably was essential to format the drive…
    RAIDxpert seemed to see it OK and I tried to do a rebuild, but it was not formatted and 30 disk errors appeared… Unfortunately because the new drive was now part of the RAID1 logical drive, these errors have been recognized by RAIDexpert as across BOTH drives – so now my source disk has 30 disk errors too..

    I tried crystaldiskinfo and it says the current sector pending count is 200 (worst 1), and the uncorrectable sector count is 200 (worst 200)…on the source drive.

    I am trying windows disk error correction tools on the C: drive, but when i try D: it says it will force a dismount, which will break the RAID status.. Any way to refresh RAIDxpert and/or reset the disk errors on D? I am cautious of starting a rebuild from a dodgy drive….

    the alternative is buy a new server and start again and try to copy the data across before the remaining drive fails completely…

    regards,
    Nigel

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.