DX Application Performance Management

Expand all | Collapse all


Jump to Best Answer
  • 1.  Heartbeat

    Posted 09-19-2014 05:13 AM



    Is it possible to get a Heartbeat alarm when a server/Agent has become gray in Introscope?

  • 2.  Re: Heartbeat

    Broadcom Employee
    Posted 09-19-2014 09:15 AM
    Not sure how we'd accomplish that since the agent is no longer reporting metrics.
    I can't even think of a way of doing that except maybe using a JS calc and a list of known agents.

  • 3.  Re: Heartbeat
    Best Answer

    Broadcom Employee
    Posted 09-19-2014 11:27 AM



    You can create an alert to monitor the metric at Custom Metric Host | Custom Metric Process | Custom Metric Agent (Collector host name or IP) | Agents | <agent host name or IP> | <agent name> | Connection Status


    1 is connected

    2 is connected but running slow

    3 is disconnected.





  • 4.  Re: Heartbeat

    Broadcom Employee
    Posted 09-19-2014 11:32 AM
    Will ConnectionStatus show 3 if the agent has already greyed out? Maybe I need more coffee this morning.

  • 5.  Re: Heartbeat

    Broadcom Employee
    Posted 09-19-2014 12:13 PM

    Hi Haruhiko,


    Yes, it will show 3.  An agent goes grey when it disconnects.


    It would be more proactive to alert on anything other than 1 because if it is connected and reporting slowly (designated by a connection status of 2) you would want to know about it.



  • 6.  Re: Heartbeat

    Broadcom Employee
    Posted 09-19-2014 12:14 PM
    See I do need more coffee. ;-)

  • 7.  Re: Heartbeat

    Broadcom Employee
    Posted 09-19-2014 12:17 PM

    I do have an IV for coffee if you need it. 

  • 8.  Re: Heartbeat

    Posted 09-22-2014 05:17 AM

    Hi Matt,

    Regarding your suggestion, it will be perfectly fine in case of standalone wily implementation.

    For clustered deployment, we face issues regarding connection status metric reporting.


    Refer - https://communities.ca.com/ideas/111998910#comment-233894171 for discussion and some useful suggestion. Hope it helps.




  • 9.  Re: Heartbeat

    Posted 10-01-2014 10:35 AM

    We do this periodically for JVMs that are prone to going "sick" but it is a bit fiddly


    Create a calculator against a java metric that should always report - e.g heap bytes in use

    Create a metric grouping from that calculator

    Create an alert from that metric grouping

    Alert if value < 1, as Heap should always have a positive value, so if is no longer reporting the calculator value will be 0


    Typically we find that these sort of jvm's will be javacoring at the same time anyway, so we have a default core file detector running via EP agent as well.