Uploaded image for project: 'OpenShift Logging'
  1. OpenShift Logging
  2. LOG-5426

[release-5.9] Cluster Logging Operator is producing stale telemetry metrics

XMLWordPrintable

    • False
    • None
    • False
    • NEW
    • VERIFIED
    • Before this update, an issue in the metrics collection code of Cluster Logging Operator caused stale telemetry metrics to be reported. With this update, a rewrite of the metrics code resolves the issue.
    • Bug Fix
    • Log Storage - Sprint 252
    • None

      Description of problem:

      The telemetry instrumentation of Cluster Logging Operator produces stale metrics containing previous and intermediate configuration. For example, creating a ClusterLogging configuration referencing the default LokiStack log storage produces an "empty" set of metrics (all options set to zero) and the "current" set of metrics:

      # Example for log_forwarder_output_info metric:
      log_forwarder_output_info{azureMonitor="0", cloudwatch="0", default="0", elasticsearch="0", fluentdForward="0", googleCloudLogging="0", http="0", kafka="0", loki="0", splunk="0", syslog="0"} 0
      log_forwarder_output_info{azureMonitor="0", cloudwatch="0", default="0", elasticsearch="0", fluentdForward="0", googleCloudLogging="0", http="0", kafka="0", loki="1", splunk="0", syslog="0"} 1

      Similarly, switching from ElasticSearch to Loki leads to the old ElasticSearch metrics still being present in the output (with a value of 0).

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

      5.9.0

      Steps to Reproduce:

      1. Install Cluster Logging Operator
      2. Configure a ClusterLogging resource
      3. Look at the metrics from the operator (prefixed with "log_")

      Actual results:

      Metric set contains stale metrics.

      Expected results:

      Metrics should reflect the current state of the configuration only.

            rojacob@redhat.com Robert Jacob
            rojacob@redhat.com Robert Jacob
            Kabir Bharti Kabir Bharti
            Votes:
            0 Vote for this issue
            Watchers:
            4 Start watching this issue

              Created:
              Updated:
              Resolved: