Uploaded image for project: 'Hawkular Metrics'
  1. Hawkular Metrics
  2. HWKMETRICS-762

Log a warning when inserting data points takes too long

    Details

    • Type: Task
    • Status: Closed (View Workflow)
    • Priority: Major
    • Resolution: Rejected
    • Affects Version/s: None
    • Fix Version/s: 0.30.1, 0.31.0
    • Component/s: REST
    • Labels:
      None

      Description

      Heapster is the primary client with which we are concerned for inserting data points. Heapster sends data points every 30 seconds by default. The interval is configurable. If Hawkular Metrics takes more than 30 seconds to complete the POST request for inserting data points, Heapster will simply drop data points for the subsequent request.

      We need to know when it takes to long to insert data because:

      1. It can cause gaps in graphs
      2. It can indicate that Hawkular Metrics and/or Cassandra cannot handle the load
      3. It can indicate that there is some other problem(s)

      The interval can be specified via a system property which can be set at start up. There is already a variable in the ansible installer, openshift_metrics_resolution, that is passed to heapster. We will assign the new system property the value of that same variable.

      There are two methods with which we are concerned:

      • CounterHandler.addData
      • GaugeHandler.addData

      These are the REST handler methods that will get called when Heapster POSTs data points.

        Gliffy Diagrams

          Attachments

            Activity

              People

              • Assignee:
                rvargasp Ruben Vargas Palma
                Reporter:
                john.sanda John Sanda
              • Votes:
                0 Vote for this issue
                Watchers:
                1 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved: