To quickly locate performance bottlenecks and troubleshoot cluster issues, enabling tracing for the cluster API Server is recommended to capture detailed request interactions. Once enabled, trace information from the API Server will be automatically sent to the Managed Service for OpenTelemetry, providing visualized trace details, real-time topology, and other monitoring data.
If you are not familiar with tracing, you can refer to Basic Concepts to understand the relevant basic concepts.
Prerequisites
You must have created an ACK managed cluster Pro edition version 1.28 or above. For upgrade instructions, see Manually Upgrade a Cluster.
The Managed Service for OpenTelemetry must be activated and authorized. For more information, see Preparations.
Managed Service for OpenTelemetry provides tools for distributed applications such as complete trace restoration, request volume statistics, trace topology, and application dependency analysis.
When using this feature, you will incur costs for the ACK cluster along with charges for the volume of observable data collected by the Managed Service for OpenTelemetry. For details on the free quota and specific pricing of the Managed Service for OpenTelemetry, refer to Billing Information.
You can configure the API Server parameters to enable tracing and set the sampling rate. After configuration, the Managed Service for OpenTelemetry console will display the detailed trace data.
Step 1: Enable tracing for the API server
Log on to the ACK console. In the left-side navigation pane, click Clusters.
On the Clusters page, find the cluster that you want to manage and click its name. In the left-side navigation pane, choose .
In the Core Components area, click Kube API Server at the bottom right of the card, then click Configure.
In the configuration panel, locate the enableTrace and samplingRatePerMillion parameters and complete the configuration as prompted on the page.
enableTrace: Enable the tracing feature for the API Server.
samplingRatePerMillion: Set the sampling rate to determine the proportion of requests sampled out of every million. Adjust the rate based on the cluster's load.
For instance, a value of 1000000 indicates a 100% sampling rate, capturing every request out of a million. A value of 100000 indicates a 10% sampling rate, while 10000 represents a 1% sampling rate.
Step 2: View trace data in OpenTelemetry
Log on to the Managed Service for OpenTelemetry console. In the left-side navigation pane, click Application List.
Select the region at the top of the page, then click the application name apiserver and follow the console guidance to view trace information.
Trace Analysis tab: Displays traces of API Server requests, application topology, and more.
For a comprehensive guide to the different tabs, refer to Application Details in the referenced document.
Enter
resources.k8s.cluster.name : "ClusterId"in the search box at the top left of the page to filter and display traces for a specific cluster.You can find the ClusterId in the cluster list.
Click a Trace ID in the Trace list to view details of a API Server request trace, including the topology view, number of services involved, and number of interfaces invoked.
Provided Services tab: Shows the number of requests, faults, and average response time for API Server API interfaces.
Dependent Services tab: Lists external components the API Server depends on, such as etcd, including the number of requests, faults, and average response time.
References
For more information on monitoring trace data for cluster data plane components, see Enable tracing for data plane components.