Finding source of TUR signals - Storage

This is a discussion on Finding source of TUR signals - Storage ; I have a fibre fabric and SAN with 8 servers (7 Dell PE 2650, 1 Dell PE 6650, XIOTech Magnitude 3D 3000e SAN, 2-McData 4400 fibre switches, a Dell PV-136T (ADIC Scalar 100) Tape Library), 4 of which are setup ...

+ Reply to Thread
Results 1 to 2 of 2

Thread: Finding source of TUR signals

  1. Finding source of TUR signals

    I have a fibre fabric and SAN with 8 servers (7 Dell PE 2650, 1 Dell PE
    6650, XIOTech Magnitude 3D 3000e SAN, 2-McData 4400 fibre switches, a Dell
    PV-136T (ADIC Scalar 100) Tape Library), 4 of which are setup in two
    2-physical node Active/Active clusters with two virtual servernames on each
    cluster. The backup system is CommVault 6.1 w/SP3. One server is Win2KAS,
    the rest are Server 2003 Enterprise SP1 or R2.

    SOMETHING in this system is generating TUR signals, and, eventually taking
    the Library and/or drives offline (after 2-3 weeks).

    I've followed the recommendations in KB 842411 in disabling the Removable
    Storage service on all servers, and putting AutoRun=0 in any ControlSet
    services that even smell like tape-related or library-related.

    But...something's still generating the TUR signal (based on the Library's
    logs), unfortunately, it doesn't tell me from what server.

    Any way to find out WHERE or what service this is coming from? Can I
    monitor each server for this signal, or something?

    Thanks



  2. Re: Finding source of TUR signals

    Did you find a fix for this problem? We are having a similar problem
    currently, again with ComVault.

    We have 2 Microsoft 2003 R2 clusters (1 x 2 node Active/Active cluster
    and 1 x 3 node Active/Active/Passive cluster) and 1 server dedicated to
    running the ComVault software and is also is dedicated to handling all
    media change operations in the library. All servers are HP DL380 with
    QLogic dual port fibre cards. The cluster nodes also have Hitachi's
    HDLM software installed to manage path failover to the HDS storage
    (AMS500)

    The tape library is a Spectra Logic T50 (2 drive LTO3, SAN connected).
    Every thing is connected to Brocade 220E switches.

    Backups operate fine - sometimes. Media change operations work fine
    initially. Then afer usually a couple of tape moves the software will
    report a media changer error and mark the tape as bad. The problem
    always occurs while trying to mount a tape.
    The errors we are getting in CommVault are:

    Failed to mount media with barcode [000087L3], side [A_70], into drive
    [IBM ULTRIUM-TD3_1], in library [SPECTRA PYTHON] on host [FILE01].
    SCSI Operation: Move Media From Slot To Drive. Reason: The SCSI device
    reported an error in the arm changer. SCSI Error - ASC[0x85],
    ASCQ[0x99]
    Advice: Please upload log files from CommServe, MediaAgent and Client
    and contact your vendor's support hotline with the job ID of the failed
    job.

    There is then a coresponding error on the library about a SCSI Robotics
    error, at this stage even moving a tape to the drive using the library
    front panel would fail. We would then have to power cycle the library
    and it will work fine.

    We have now had 3 replacement libraries from Spectra, along with
    firmware upgrades etc. We have also applied KB842411 (disabling AutoRun
    etc) - but to no avail.

    The Unknown Media Changer on all the cluster nodes has also been
    disabled, so that only the main backup server can control the library
    (as is setup in the backup software) - therefore hopefully stoping
    Windows even trying to access the media changer.

    We are now getting to the stage where we are looking at other
    manufacturers for the tape library (HP MSL4048 for example)

    Any thoughts/help would be appreciated.

    Bill Bradley wrote:
    > I have a fibre fabric and SAN with 8 servers (7 Dell PE 2650, 1 Dell PE
    > 6650, XIOTech Magnitude 3D 3000e SAN, 2-McData 4400 fibre switches, a Dell
    > PV-136T (ADIC Scalar 100) Tape Library), 4 of which are setup in two
    > 2-physical node Active/Active clusters with two virtual servernames on each
    > cluster. The backup system is CommVault 6.1 w/SP3. One server is Win2KAS,
    > the rest are Server 2003 Enterprise SP1 or R2.
    >
    > SOMETHING in this system is generating TUR signals, and, eventually taking
    > the Library and/or drives offline (after 2-3 weeks).
    >
    > I've followed the recommendations in KB 842411 in disabling the Removable
    > Storage service on all servers, and putting AutoRun=0 in any ControlSet
    > services that even smell like tape-related or library-related.
    >
    > But...something's still generating the TUR signal (based on the Library's
    > logs), unfortunately, it doesn't tell me from what server.
    >
    > Any way to find out WHERE or what service this is coming from? Can I
    > monitor each server for this signal, or something?
    >
    > Thanks



+ Reply to Thread