1 Reply Latest reply on Oct 14, 2011 8:13 PM by rgward

    Additional Granularity in Discovery Scheduling

    mdriskell

      Now I know that in 10.2 the discovery process should be sped up but I still think there could be some additional control.

       

      I would like to have a day of the week type scheduling ability.

      Example  

      172.16.0.0/16 Monday 12:00 AM

      172.17.0.0/16 Tuesday & Thursday 12:00 AM

      172.18.0.0/16 Wednesday 12:00 AM

      You get the idea...But extra granularity would be very handy.  Our alerts allow the ability to only execute on certain days as does the report scheduler why not discoveries.

      The only way I now to do this now would be to setup the discovery I want 172.16.0.0/16 set it to custom 168 hours and run it at 12:00 AM on monday to get it to follow that schedule which is far less than ideal (I'm not even certain this would work)

        • Re: Additional Granularity in Discovery Scheduling
          rgward

          +1 here.  Definitely would like to see day of the week type scheduling ability.

          Also have the ability to specify a discovery window, for example 6pm - 6am.  If the discovery is not complete by 6am, it would stop and post the discovered results and status where the discovery stopped (last address).  As it works today, when the discovery timeout (6 hours) is exceeded it throws an error and all discovered nodes are lost.  You have go back and tweak your discovery criteria and try again.

          Plus, support for concurrent discoveries.  Today, all discoveries are run on the primary engine.  One option would be to add functionality to Network Discovery to schedule discoveries on additional polling engines.   So, if Network Discovery could do at least 2 concurrent discoveries and you have 2 polling engines then 4 concurrent discoveries would be possible.  That would be sweet! Single threaded discoveries just doesn't cut it for a large organization like us which supports several class B networks.