Emergency

Feedback Needed: Experimental CPU Details dashboard for PMM

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Feedback Needed: Experimental CPU Details dashboard for PMM

    Click image for larger version

Name:	
Views:	1
Size:	153.3 KB
ID:	50375 I have created experimental dashboard for possible inclusion in later Percona Monitoring and Management version, and I though it might be great to get some feedback about it before including it in release

    Dashboard is available from Grafana.com so it should be very easy to add it to your PMM Installation

    https://grafana.com/dashboards/4151

    Please let me know if there is anything to improve before we include it in release.

    Couple of things to note
    • Note red dots in CPU overview graph. It shows total CPU usage for busiest core. Very often we see some cores completely saturated being a bottleneck while overall CPU usage looks low
    • The table shows utilization details for all the CPU cores to help to spot the problems, for example if one core has high IOwait it will likely to correspond to single thread IO intensive process being blocked by IO a lot.
    Dashboard Example

  • #2
    Forgive me if this is already part of the design.

    The idea of setting baselines,
    and then setting yellow (hey!), orange (WARNING) and red (ERROR) thresholds appeals to me as a generic sensibility for most monitoring.

    Comment


    • #3
      Thanks Jeremy,

      It would be cool through it does not do that due to platform (Grafana) limitations - I can only set static colors for SingleStat panel type used here.

      Comment

      Working...
      X