DX NetOps

Expand all | Collapse all

Tuesday Tip: Can a running or hanging SPECTRUM AutoDiscovery be aborted?

  • 1.  Tuesday Tip: Can a running or hanging SPECTRUM AutoDiscovery be aborted?

    Broadcom Employee
    Posted 12-02-2011 02:46 AM

    Tuesday Tip: Can a running or hanging SPECTRUM AutoDiscovery be aborted?

    CA Spectrum Tuesday Tip by Bill Wolforth, Support Engineer, November 30, 2011

    How to abort a running or hanging SPECTRUM AutoDiscovery session.
    Description: There may be times when there is a need to abort a running or hanging AutoDiscovery. In these instances the following processes can be used.

    Solution :

    OneClick Option:

    In the OneClick Topology view, select the VNM icon, then in the Component Details Pane access the Information Tab then expand the AutoDiscovery Control selection. Click on the "Abort Discovery" button.

    Command Line Interface (CLI) Option:

    The CLI Action code to abort and AutoDiscovery is 0x25e004d. This action code must be sent to the model of type TopologyWrkSpc (Model Type Handle =0x0x25e0001).

    show models mth=0x25e0001
    MHandle MName MTypeHnd MTypeName
    0x200002d AutoDiscovery 0x25e0001 TopologyWrkSpc

    update action=0x25e004d mh=0x200002d

    NOTE: The model handle will be different than what is used in the above example.



  • 2.  RE: Tuesday Tip: Can a running or hanging SPECTRUM AutoDiscovery be aborted

    Posted 12-12-2011 11:25 AM
    Hi Bill!

    Thanks for this great Tuesday Tip on how to abort a running or hanging SPECTRUM AutoDiscovery session.

    Cheers!
    Mary


  • 3.  Re: Tuesday Tip: Can a running or hanging SPECTRUM AutoDiscovery be aborted?

    Broadcom Employee
    Posted 08-15-2018 07:39 AM

    CA Spectrum AutoDiscovery job-queue could be managed by actions send to the TopologyWrkSpc model:

     

    ./update action=0x10601 mh=<TopologyWrkspc> == show AutoDiscovery queue (single threaded)
    ./update action=0x10602 mh=<TopologyWrkspc> == kill AutoDiscovery "current job"

     

    Once an ADisc job is canceled (dropped) this will report message to the ./SS/VNM.OUT file - covering:

    Clearing the pending rediscovery queue @<date> <time>

    (sample:  Clearing the pending rediscovery queue @2018/07/23 19:17:34 )

     

    At SpectroSERVER level the ADisc job queue is reported by specific event: - thus any "new queue discovery action" will create event of type 0x25e0004 / 0x25e0006 -> "Auto-discovery mapping process can not run on model "XXXXX" of type "YYYYY" because another mapping process is currently running" (0004)  - so it is queued(0006). Doing then action=0x10601 and you see/find increased ADisc queue. 

     

    By default the AutoDiscovery job-queue is worked single_threaded to ensure proper modeling.

     

    hth