none
HCK Test Code Fixes RRS feed

  • Question

  • Is there any public documentation available that provides "code fix" information when new versions of the HCK are released?  After working through a multitude of failures with the MPIO test and iSCSI a new failure has popped up.   Even though the test is running iSCSI the failure states;

    (x) For non-iSCSI test environment, please make sure at least two HBA ports are connected to the storage target.

    Since this is the first time we have seen this I am wondering if a code change in the MPIO test may have caused this.

    Thanks,

    Monday, December 3, 2012 10:49 PM

Answers

All replies

  • For HCK QFE release note, please refer below

    http://msdn.microsoft.com/zh-tw/windows/hardware/jj553782

    Thanks, Steve

    Tuesday, December 4, 2012 1:55 AM
  • If using iSCSI, how many connections are linked to the target? Are you using multiple sessions instead of multiple physical IP addresses?

    Friday, December 7, 2012 6:35 PM
  • There are two connections linked to the target.   One NIC to controller 0 and one NIC to controller 1, our RAID Array does not support mutliple physical IP addresses through a single port with iSCSI.  This is due to the use of the MS iSCSI initiator being the host side target.

    The test log file error states:

    (X) For non-iSCSI test environment, please make sure at least two HBA ports are connected to the storage target.

    The Storage_log.xml lists the bus type as iSCSI.

      <StorageDevice>

      <DeviceID>ROOT\MPIO\0002</DeviceID>

      <FriendlyName />

      <StorageClass>MPIO DSM</StorageClass>

      <BusType>BusTypeiScsi</BusType>

      <MPIODevice />

      </StorageDevice>

    Monday, December 10, 2012 6:36 PM
  • Could you configure like this?

    Setup one connection to the target

    Create two sessions.

    Then run the MPIO test. Let's see what the result is.

    Wednesday, January 30, 2013 11:37 PM
  • Our array does not support two sessions through a single connection for iSCSI.  This is due to the iSCSI initiator being the target on the host side and only having one IQN. 

    I received the following feedback through another channel and it does fix the problem although not an ideal solution.

    The configuration information of the client computer is gathered during the installation of the HCK client software.  When the client computer configuration is changed, the HCK client software should be un-installed and, as a further precaution, the name of the computer changed as well.  There were some instances where the gatherer information of the same named computers were corrupted due to sharing the same name.  The newly re-named computer can be added to the same machine pool for testing.

    Thank you,

    Dawn

    Monday, February 4, 2013 11:24 PM