This topic describes how to use the global check feature to troubleshoot issues in Container Service for Kubernetes (ACK) clusters and networks.

Prerequisites

  • An ACK cluster is created. For more information, see Create an ACK managed cluster.
  • The cluster is running as expected. You can navigate to the Clusters page and check whether the cluster is in the Running state.

Scenarios

To perform a global check on your cluster resources, components, and configurations and obtain suggestions on how to fix issues, you need only to click Start. No parameter configurations are required. We recommend that you check and maintain your cluster on a regular basis.

Procedure

  1. Log on to the ACK console.
  2. In the left-side navigation pane of the ACK console, click Clusters.
  3. On the Clusters page, find the cluster on which you want to perform a check and choose More > Cluster Check in the Actions column.
  4. In the left-side navigation pane of Container Intelligence Service, choose Cluster Check > Cluster Check. You can also click Cluster Check in the Toolbox section to go to the Cluster Check page.
    Cluster Check
    • Cluster: displays the cluster name, type, status, and the most recent check result.
    • Cluster check: displays the time and results of the previous five checks in reverse chronological order. If this is the first time that you perform a cluster check, the Start button appears in this section. You can click Start to go to the Cluster Check page.
    • Toolbox: displays features that are related to cluster health checks. The cluster check feature is available. More features such as node health checks will be available soon.
  5. Click Start. In the Cluster Check panel, read and select the Warning check box, and click Start.
  6. To view the progress of the global check, click Details. On the Report page, you can view the status of cluster resources and cluster configurations.
    Report
  7. Perform an action based on the check result.
    • Normal: No actions are required.
    • Warning: Confirm the severity of each issue. Troubleshoot the issues that may cause cluster anomalies.
    • Error: Troubleshoot the issues at the earliest opportunity to avoid cluster errors.

    You can click the color box to view instance information and the detailed check result. The check result includes check items, possible impacts, and suggestions on how to fix issues. The information helps you troubleshoot issues.

Other checks

  • Upgrade check: the precheck before an ACK cluster is upgraded.
  • Component check: the precheck before a component is upgraded.
  • Migration check: the precheck before an ACK dedicated cluster or standard cluster is migrated to an ACK Pro cluster.
  • Scheduled inspection: the scheduled cluster check.
  • Cluster diagnosis: If you cannot locate the causes of node, pod, and network issues based on the check report, we recommend that you use the cluster diagnosis feature to troubleshoot the issues. For more information, see Use the cluster diagnosis feature to troubleshoot cluster issues.