Topic
  • 5 replies
  • Latest Post - ‏2013-06-23T18:44:03Z by blamberty
blamberty
blamberty
4 Posts

Pinned topic hot spare difficulty

‏2013-05-21T01:30:38Z |

We have a x3500 system running Windows Server 2003 with ServeRAID 8k controller and two RAID 5 logical drives with 3 physical drives each.  After adding another hot swap drive identical to one we used to replace a failed drive (same FRU 39R7348), we can't get the hot spare feature on this 7th physical drive to work fully.  In ServeRAID Manager after selecting the option "Create hot-spare drive" for this drive, the yellow plus sign appears next to this drive icon, indicating that it is too small, or some other problem.  We know this setup supported an active hot spare, since we used to have a 7th drive acting as a hot spare, and the RAID 5 logical drives had status Yes on the Protected by hot-spare property.  Now the status is No on this property.  Reinitializing the drive does not seem to help. When we view the physical drive list in Text description view, all 7 drives report a size of 68.37 GB.  What remedy do you suggest?  Thanks.

  • Novikov_Alexander
    Novikov_Alexander
    6877 Posts

    Re: hot spare difficulty

    ‏2013-05-21T13:34:57Z  

    Dear Brett,

    could you provide the DSA log of your server.

    Best regards,
    Alexander Novikov
    Russia, Moscow

  • blamberty
    blamberty
    4 Posts

    Re: hot spare difficulty

    ‏2013-05-21T17:35:24Z  

    Dear Alexander,

    I just ran the portable DSA utility.  I have attached a ZIP file with all the files DSA created.

    Thanks.

    Brett

    Attachments

  • Novikov_Alexander
    Novikov_Alexander
    6877 Posts

    Re: hot spare difficulty

    ‏2013-05-21T19:04:50Z  
    • blamberty
    • ‏2013-05-21T17:35:24Z

    Dear Alexander,

    I just ran the portable DSA utility.  I have attached a ZIP file with all the files DSA created.

    Thanks.

    Brett

    Dear Brett,

    first of all update server's software:

        IBM BIOS Flash Update v1.74
        IBM System x3500 Baseboard Management Controller (BMC) v1.42
        IBM ServeRAID 8k BIOS and Firmware v5.2-0-17003

        Mapping Layer for OSA IPMI on x86 v1.18
        ATI RADEON Video Driver v6.14.10.6744
        Broadcom NetXtreme Device Driver for Windows v11.7.0.0
        IBM ServeRAID 8k Controller Driver v5.2.0.12913
        IBM ServeRAID Manager 9.30

    Best regards,
    Alexander Novikov
    Russia, Moscow

  • blamberty
    blamberty
    4 Posts

    Re: hot spare difficulty

    ‏2013-05-23T16:04:56Z  

    Dear Alexander,

    Thanks for the advice on updating server software.  I will do so when I can take down the server in the next two weeks.

    In the meantime, I have more information about the offending hard drive that may help solve the problem.  All 6 drives in the two RAID 5 logical drives report the following in the capacity list:

    Segment 0 Blocks: 143,155,200 (RAID Level 5)

    Segment 1 Blocks: 128 (Metadata)

    Segment 2 Blocks: 218,672 (Metadata)

    The offending spare drive reports:

    Segment 0 Blocks: 128 (Metadata)

    Segment 1 Blocks: 218,800 (Metadata)

    Segment 2 Blocks: 143,155,072 (Spare)

     

    It looks like 128 more blocks are taken for Metadata leaving 128 less blocks for the spare, and thus this is less than the capacity of the other drives.  I tried reinitializing the drive, but that did not change the numbers.  Is there some other operation I can perform to make more blocks available to the spare?  Or is this something that might be cured by all the software updates?

    Thanks again,

    Brett

  • blamberty
    blamberty
    4 Posts

    Re: hot spare difficulty - resolved

    ‏2013-06-23T18:44:03Z  
    • blamberty
    • ‏2013-05-23T16:04:56Z

    Dear Alexander,

    Thanks for the advice on updating server software.  I will do so when I can take down the server in the next two weeks.

    In the meantime, I have more information about the offending hard drive that may help solve the problem.  All 6 drives in the two RAID 5 logical drives report the following in the capacity list:

    Segment 0 Blocks: 143,155,200 (RAID Level 5)

    Segment 1 Blocks: 128 (Metadata)

    Segment 2 Blocks: 218,672 (Metadata)

    The offending spare drive reports:

    Segment 0 Blocks: 128 (Metadata)

    Segment 1 Blocks: 218,800 (Metadata)

    Segment 2 Blocks: 143,155,072 (Spare)

     

    It looks like 128 more blocks are taken for Metadata leaving 128 less blocks for the spare, and thus this is less than the capacity of the other drives.  I tried reinitializing the drive, but that did not change the numbers.  Is there some other operation I can perform to make more blocks available to the spare?  Or is this something that might be cured by all the software updates?

    Thanks again,

    Brett

    The problem was resolved by setting the drive to defunct, removing it, and reinserting it.  It initialized correctly this time, with capacity available matching the other drives.  Now it functions as a hot spare with the blue plus sign.  The problem may have arisen that it was originally inserted along with the other replacement drive and was not correctly initialized while the other drive was being rebuilt.

     

    Thanks,

    Brett