Container Service for Kubernetes (ACK) integrates with Managed Service for Prometheus and Managed Service for Grafana, so you do not need to build your own monitoring system or visualization dashboard. After you deploy applications in Knative, you can ingest monitoring data into Prometheus and view Knative performance metrics in Grafana.
Prerequisites
-
Knative is deployed in your cluster. For more information, see Deploy and manage Knative components.
-
Managed Service for Prometheus is enabled for your ACK cluster. For more information, see Configure Managed Service for Prometheus (recommended) or Open source Prometheus monitoring.
Billing
After you ingest the monitoring data of your Knative services into Managed Service for Prometheus, the relevant add-ons automatically send monitoring metrics to the service. These metrics are treated as custom metrics. Using custom metrics incurs additional fees.
To avoid additional fees, before you enable this feature, read the Billing overview of Managed Service for Prometheus to understand how custom metrics are billed. Fees vary based on factors such as your cluster scale and the number of applications. You can use the resource usage statistics feature to monitor and manage your resource usage.
Enable Managed Service for Prometheus monitoring
-
Log on to the ACK console. In the left navigation pane, click Clusters.
-
On the Clusters page, click the name of your cluster. In the left navigation pane, click .
-
On the Knative page, click the Monitoring Dashboards tab and follow the on-screen instructions to enable the Prometheus dashboard.
View the Knative monitoring dashboard
After you enable Managed Service for Prometheus monitoring, you can view detailed monitoring data on the Monitoring Dashboards tab.
|
Category |
Description |
|
Request data |
In the Overview (average over the selected time range) section, view the following Knative metrics:
The vertical axis (ops/sec) of the Request Volume, 4xx, and 5xx charts indicates the number of requests processed per second.
|
|
Response latency data |
In the Response Time section, view Knative response latency data, including P50, P90, P95, and P99.
|
|
Request concurrency |
In the Autoscaler section, view detailed request concurrency data for Knative.
|
|
Activator |
In the Activator section, view the following metrics for the Knative Activator:
|
|
Resource usage |
In the Resource Usages section, view Knative resource usage, including CPU and memory.
|
View key Knative service discovery metrics
You can view and verify the monitoring data of Knative services by selecting metrics or writing PromQL queries.
-
Log on to the Managed Service for Prometheus console. In the left-side navigation pane, click Integration Management.
-
On the Integrated Environments page, click the Container Service tab. Then, in the Integrated Addons column, click Knative.
-
Click the Metrics Explorer tab to view the following key metrics for Knative.
For information about the UI workflow, see Metrics Explorer.
Note-
If a Knative application scales to zero and has no pods, Managed Service for Prometheus cannot collect pod metrics such as request concurrency and requests per second. The console displays these metrics only after you access the pods of the Knative application.
-
If a Knative application is not scaled to zero, the console displays pod metrics such as request concurrency and requests per second without requiring you to access the pods.
Metric
Description
queue_average_concurrent_requests
The request concurrency of a pod.
queue_requests_per_second
The number of requests per second for a pod.
revision_request_latencies
The response latency.
revision_request_count
The request count of a Revision.
-
References
For information about how to use Simple Log Service (SLS) for log collection, monitoring, and alerting, see Collect Logs from Knative Services and Configure alerting for Knative Services.