Backup Exec 8.6 Cluster Failover - Veritas Backup Exec

This is a discussion on Backup Exec 8.6 Cluster Failover - Veritas Backup Exec ; I have a W2k Cluster running Bexec 8.6, the tape library is a Compaq SSL2020 connected via a MDR into a switch then to a 176479-B21 (Emulex LP8000, KGPSA) The Bexce is installed in clustered mode and works fine, Until ...

+ Reply to Thread
Results 1 to 2 of 2

Thread: Backup Exec 8.6 Cluster Failover

  1. Backup Exec 8.6 Cluster Failover


    I have a W2k Cluster running Bexec 8.6, the tape library is a Compaq SSL2020
    connected via a MDR into a switch then to a 176479-B21 (Emulex LP8000, KGPSA)
    The Bexce is installed in clustered mode and works fine, Until I failover,
    then all the tape devices appear offline on the new node. Its W2K SP2 and
    the latest Build of Bexec, I have two clusters both with MDR's and SSL2020's,
    both do the same.

    Any help gratefully received.

    Andy

  2. Re: Backup Exec 8.6 Cluster Failover

    Andy,

    The problem you describe is common in the equipment you have if left with
    everything at factory defaults.

    I am assuming the you also have the Compaq OEM Brocade switch, if so, the
    firmware by default is 2.17, in some cases Compaq upgraded to 2.19m. This
    is still not high enough to resolve the bus reset issue when using Emulex
    cards either Standard or OEM. You must upgrade your firmware above 2.21a.
    However, Compaq doesn't support anything between 2.19m and 2.6 so you will
    have to go to the latest which is available on their website.

    The other thing that you must do is modify your LP8000 to enable bus reset
    and verify that you have the latest driver and firmware from either Compaq
    or Emulex. (You must be sure that you get the right driver and firmware
    combination so be absolutely sure which card you have) Contact Compaq or
    Emulex for the proper settings for your equipment.

    The reason the problem occurs is that SCSI-2 reservations are used to
    acquire control of the shared tape device or loader. When the failover
    occurs, the node that previously had control has not relenquished it so that
    the second node cannot connect to the devices. If a bus reset could occur
    when the services for BE were started, then BE could grab the drives and
    everything would be fine. By following the above steps you should be able
    to simply resolve your issue. It is not Backup Exec related, any backup
    software in your hardware configuration would encounter the same problem.

    Hope this helps,

    -kk


    "Andy" wrote in message news:3d3ee924$1@hronntp01....
    >
    > I have a W2k Cluster running Bexec 8.6, the tape library is a Compaq

    SSL2020
    > connected via a MDR into a switch then to a 176479-B21 (Emulex LP8000,

    KGPSA)
    > The Bexce is installed in clustered mode and works fine, Until I failover,
    > then all the tape devices appear offline on the new node. Its W2K SP2 and
    > the latest Build of Bexec, I have two clusters both with MDR's and

    SSL2020's,
    > both do the same.
    >
    > Any help gratefully received.
    >
    > Andy




+ Reply to Thread