×

Monitoring the overall health status of Knative components

You can use the OpenShift Container Platform monitoring dashboards to view the overall health status of Knative.

Prerequisites
  • You have cluster administrator permissions, and access to the Administrator perspective in the OpenShift Container Platform web console.

  • You have installed the OpenShift Serverless Operator, as well as the Knative Serving or Knative Eventing components.

  • The OpenShift Container Platform monitoring stack is enabled on your cluster. You can enable monitoring for OpenShift Serverless during installation by checking the box to Enable operator recommended cluster monitoring on this namespace when installing the OpenShift Serverless Operator.

Procedure
  1. In the Administrator perspective, navigate to MonitoringDashboards.

  2. Select the Knative Health Status dashboard in the Dashboard drop-down to view the overall health status of Knative. If your Knative deployment is running as expected, the dashboard shows a status of Ready.

    Knative Health Status dashboard

    If you have Knative Serving or Knative Eventing installed, you can also scroll down to see the health status for each of these components.

Monitoring Knative Serving revision CPU and memory usage

You can use the OpenShift Container Platform monitoring dashboards to view revision CPU and memory usage metrics for Knative Serving components.

Prerequisites
  • You have cluster administrator permissions, and access to the Administrator perspective in the OpenShift Container Platform web console.

  • You have installed the OpenShift Serverless Operator, as well as the Knative Serving component.

  • The OpenShift Container Platform monitoring stack is enabled on your cluster. You can enable monitoring for OpenShift Serverless during installation by checking the box to Enable operator recommended cluster monitoring on this namespace when installing the OpenShift Serverless Operator.

Procedure
  1. In the Administrator perspective, navigate to MonitoringDashboards.

  2. Select the Knative Serving - Source CPU and Memory Usage dashboard in the Dashboard drop-down list to view the following metrics:

    • Total CPU Usage (rate per minute)

    • Total Memory Usage (bytes)

    • Total Network I/O (rate per minute)

    • Total Network Errors (rate per minute)

  3. Optional: You can filter this dashboard by Namespace, Configuration, or Revision, by selecting an option from the drop-down list.

Monitoring Knative Eventing source CPU and memory usage

You can use the OpenShift Container Platform monitoring dashboards to view source CPU and memory usage metrics for Knative Eventing components.

Prerequisites
  • You have cluster administrator permissions, and access to the Administrator perspective in the OpenShift Container Platform web console.

  • You have installed the OpenShift Serverless Operator, as well as the Knative Eventing component.

  • The OpenShift Container Platform monitoring stack is enabled on your cluster. You can enable monitoring for OpenShift Serverless during installation by checking the box to Enable operator recommended cluster monitoring on this namespace when installing the OpenShift Serverless Operator.

Procedure
  1. In the Administrator perspective, navigate to MonitoringDashboards.

  2. Select the Knative Eventing - Source CPU and Memory Usage dashboard in the Dashboard drop-down list to view the following metrics:

    • Total CPU Usage (rate per minute)

    • Total Memory Usage (bytes)

    • Total Network I/O (rate per minute)

    • Total Network Errors (rate per minute)

Monitoring event sources

You can use the OpenShift Container Platform monitoring dashboards to view metrics for event sources in your cluster.

Prerequisites
  • You have cluster administrator permissions, and access to the Administrator perspective in the OpenShift Container Platform web console.

  • You have installed the OpenShift Serverless Operator, as well as the Knative Eventing component.

  • The OpenShift Container Platform monitoring stack is enabled on your cluster. You can enable monitoring for OpenShift Serverless during installation by checking the box to Enable operator recommended cluster monitoring on this namespace when installing the OpenShift Serverless Operator.

Procedure
  1. In the Administrator perspective, navigate to MonitoringDashboards.

  2. Select the Knative Eventing - Sources dashboard in the Dashboard drop-down list.

  3. You can now view the following metrics:

    1. For API server sources:

      • Event Count (rate per minute)

      • Success Rate (2xx Event, fraction rate per minute)

      • Event Count by Response Code Class (rate per minute)

      • Failure Rate (non-2xx Event, fraction rate per minute)

    2. For ping sources:

      • Event Count (rate per minute)

      • Success Rate (2xx Event, fraction rate per minute)

      • Event Count by Response Code Class (rate per minute)

      • Failure Rate (non-2xx Event, fraction rate per minute)

    3. For Kafka sources:

      • Event Count (rate per minute)

      • Success Rate (2xx Event, fraction rate per minute)

      • Event Count by Response Code Class (rate per minute)

      • Failure Rate (non-2xx Event, fraction rate per minute)

Monitoring Knative Eventing brokers and triggers

You can use the OpenShift Container Platform monitoring dashboards to view metrics for brokers and triggers in your cluster.

Prerequisites
  • You have cluster administrator permissions, and access to the Administrator perspective in the OpenShift Container Platform web console.

  • You have installed the OpenShift Serverless Operator, as well as the Knative Eventing component.

  • The OpenShift Container Platform monitoring stack is enabled on your cluster. You can enable monitoring for OpenShift Serverless during installation by checking the box to Enable operator recommended cluster monitoring on this namespace when installing the OpenShift Serverless Operator.

Procedure
  1. In the Administrator perspective, navigate to MonitoringDashboards.

  2. Select the Knative Eventing - Broker/Trigger dashboard in the Dashboard drop-down list.

  3. You can now view the following metrics:

    1. For brokers:

      • Event Count (avg/sec, over 1m window)

      • Success Rate (2xx Event, fraction rate, over 1m window)

      • Event Count by Event Type (avg/sec, over 1m window)

      • Event Count by Response Code Class (avg/sec, over 1m window)

      • Failure Rate (non-2xx Event, fraction rate, over 1m window)

      • Event Dispatch Latency (ms)

    2. For triggers:

      • Event Count (avg/sec, over 1m window)

      • Success Rate (2xx Event, fraction rate, over 1m window)

      • Event Count by Event Type (avg/sec, over 1m window)

      • Event Count by Response Code Class (avg/sec, over 1m window)

      • Failure Rate (non-2xx Event, fraction rate, over 1m window)

      • Event Dispatch Latency (ms)

      • Event Processing Latency (ms)

Monitoring Knative Eventing channels

You can use the OpenShift Container Platform monitoring dashboards to view metrics for channels in your cluster.

Prerequisites
  • You have cluster administrator permissions, and access to the Administrator perspective in the OpenShift Container Platform web console.

  • You installed the OpenShift Serverless Operator, the Knative Eventing component, and the KnativeKafka custom resource.

  • The OpenShift Container Platform monitoring stack is enabled on your cluster. You can enable monitoring for OpenShift Serverless during installation by checking the box to Enable operator recommended cluster monitoring on this namespace when installing the OpenShift Serverless Operator.

Procedure
  1. In the Administrator perspective, navigate to MonitoringDashboards.

  2. Select the Knative Eventing - Channel dashboard in the Dashboard drop-down list.

  3. You can now view the following metrics:

    1. For in-memory channels:

      • Event Count (avg/sec, over 1m window)

      • Success Rate (2xx Event, fraction rate, over 1m window)

      • Event Count by Response Code Class (avg/sec, over 1m window)

      • Failure Rate (non-2xx Event, fraction rate, over 1m window)

      • Event Dispatch Latency (ms)

    2. For Kafka channels:

      • Event Count (avg/sec, over 1m window)

      • Success Rate (2xx Event, fraction rate, over 1m window)

      • Event Count by Response Code Class (avg/sec, over 1m window)

      • Failure Rate (non-2xx Event, fraction rate, over 1m window)

      • Event Dispatch Latency (ms)