Uploaded image for project: 'OpenShift Logging'
  1. OpenShift Logging
  2. LOG-4689

[release-5.5] Show FluentD Buffer Usage in metrics dashboard instead of availability

    XMLWordPrintable

Details

    • False
    • None
    • False
    • NEW
    • VERIFIED
    • Hide
      Before this update, an error in the query used for the "FluentD Buffer Availability" graph in the metrics dashboard created by the Cluster Logging Operator caused the minimum buffer usage to be displayed instead. With this update, the graph has been renamed to "FluentD Buffer Usage" and the query has been fixed to show the maximum buffer usage.
      Show
      Before this update, an error in the query used for the "FluentD Buffer Availability" graph in the metrics dashboard created by the Cluster Logging Operator caused the minimum buffer usage to be displayed instead. With this update, the graph has been renamed to "FluentD Buffer Usage" and the query has been fixed to show the maximum buffer usage.
    • Bug Fix

    Description

      Description of problem:

      This PR:

      • fixes that "FluentD Buffer Availability" can show us the buffer usage based on the most highest usage value.

      If "fluentd_output_status_buffer_available_space_ratio" metrics returns 100 and 50, "max" aggregation operator(current) always show us 0%(100-100) buffer usage, not 50%(100-50) buffer usage. It's not helpful to detect collector pods in high buffer usage. So it should change it to "min" aggregation operator for showing us values based on the most highest usage of buffer.

      Version-Release number of selected component (if applicable):

      OpenShift Logging 5.7+

      How reproducible:

      Always

      Steps to Reproduce:

      1. Install OpenShift Logging 5.7+
      2. Observe → Dashboards in web console
      3. On the Dashboards page, select OpenShift Logging from the Dashboard menu
      4. You can see "FluentD Buffer Availability" graph there

      Actual results:

      "FluentD Buffer Availability" show us usage of buffer based on "max" aggregator operator. As a result, it shows us the lower buffer usage than real usages.
      And "FluedD Buffer Availability" shows us usage, not availalable value of the buffer. It's not matched with the title of graph.

      Expected results:

      • Title change: "FluedD Buffer Availability" -> "FluentD Buffer Usage"
      • Fixes the promQL of the graph as follows.
      100 - min by(plugin_id)(fluentd_output_status_buffer_available_space_ratio
      

      Additional info:

      Filed PR is here: https://github.com/openshift/cluster-logging-operator/pull/2172

      Attachments

        Issue Links

          Activity

            People

              rojacob@redhat.com Robert Jacob
              rhn-support-dapark Daein Park
              Kabir Bharti Kabir Bharti
              Votes:
              0 Vote for this issue
              Watchers:
              1 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: