DX NetOps

  • 1.  Average, Maximum,Minimum

    Posted Aug 17, 2017 05:16 PM

    Can someone help me better understand how the Average, maximum and minimum values for any metric are calculated in PM. So is that based off of poll rate? But how does PM assume an average when you already poll for example a percentage? We are trying to compare the values from a router/switch command line result with PM. These are two very different figures.



  • 2.  Re: Average, Maximum,Minimum
    Best Answer

    Broadcom Employee
    Posted Aug 18, 2017 02:30 AM

    Please see the wiki article here: Minimum and Maximum Values - CA Performance Management - 3.2 - CA Technologies Documentation 

    If the current view resolution is as-polled/5min, average, min and max reflect the polled value. The idea is to have consistent attribute names, independent from the resolution setting.



  • 3.  Re: Average, Maximum,Minimum

    Posted Aug 18, 2017 10:06 AM

    I totally understand the as-polled/5min resolution as being the actual polled value for no matter what the min/max/avg is. But after reading the doc, it left me scratching my head. It speaks of min and max. But I assume the average is taken of all the values and that is how the rollup is calculated.



  • 4.  Re: Average, Maximum,Minimum

    Broadcom Employee
    Posted Aug 18, 2017 10:24 AM

    yes, correct, If you think the wiki article should clarify also the average calculation, please feel free to leave a comment on that page.



  • 5.  Re: Average, Maximum,Minimum

    Broadcom Employee
    Posted Aug 22, 2017 08:46 AM

    Hi, I will clarify the average calculation and add these details to the documentation.



  • 6.  Re: Average, Maximum,Minimum

    Broadcom Employee
    Posted Aug 24, 2017 10:32 AM

    Updated details including new information about the Total calculations are now available here: Total, Average, Minimum, and Maximum Values - CA Performance Management - 3.2 - CA Technologies Documentation 



  • 7.  Re: Average, Maximum,Minimum

    Posted Aug 24, 2017 10:38 AM

    Thank You!