DX Application Performance Management

Expand all | Collapse all

Feature Request - Maximum Response Time Metric

  • 1.  Feature Request - Maximum Response Time Metric

    Posted 09-26-2018 08:47 PM

    Anyone in favor of a feature request for CA APM to include a Maximum Response Time metric column in URL and SQL "Overview" summary tables in addition to the average response time metric?   We deal with many intermittent scenarios where a transaction encounters a response time spike (1 second spiking to 15 seconds).  CA APM doesn't provide an easy means to identify the specific offending transaction that encountered an intermittent spike.

     

    The average response time metric is not indicative of a transaction which encounters infrequent intermittent spikes.  The MAX ART metric would solve this.   The present CA APM workflow requires the end-user to click on each transaction and view the trending average response time metrics with the "Max/Min" setting checked which is an extremely poor end-user experience.

     

    Anyone in favor of the MAX ART metric?  Would also be good to have alerting capabilities on MAX ART metric.



  • 2.  Re: Feature Request - Maximum Response Time Metric

    Posted 09-26-2018 09:16 PM

    Hi Jeff,

    I converted your thread to a discussion but probably the best option is to submit it as an Idea and then other community users can comment there and vote on it (under the pencil icon drop-down or Actions drop-down menus you will see the Idea option)

    Thanks

    Lynn



  • 3.  Re: Feature Request - Maximum Response Time Metric

    Posted 09-27-2018 08:59 AM

    Thanks Lynn.   I'm a newbie to the CA Communities



  • 4.  Re: Feature Request - Maximum Response Time Metric

    Posted 10-07-2018 05:51 AM

    Hello Jeff,

    you are right, this is one of the CA APM problems. It is lacking in some more useful/individual ways to analyse the collected metrics data. The metrics data is available, but no way to analyse the metrics data.

    This is one reason you find a hint to Grafana (see APMSQL Cookbook - CA Application Performance Management - 10.7 - CA Technologies Documentation), I think.

     

    I copy the "importent" metrics data into a Postgres DB to analyse the collected metrics data. This is awkward to do, but more easy to handle and I can group the metrics data based on the used queries to select the data, without creating additional matrics data at the collection time.

     

    See Using Grafana with CA APMSQL - CA Application Performance Management - 10.7 - CA Technologies Documentation, this is a suggestion only.

     

    With regards,

    Lutz

     

    p.s.

    APM 10.5+ (APMSQL must be available) -> APM 10.7+ for AIX APMSQL
    Tomcat 8+ -> Jetty is supported also
    Grafana 4.6.0+ -> Linux only