TKE Monitoring

Last updated: 2021-02-03 14:28:25

    New TKE Monitoring Features

    ● Monitored objects can be updated automatically.
    ● Workload/Component/Node monitoring scenarios are added.
    ● More monitoring metrics are added. The total number of new TKE metrics reaches 140.
    ● You can block a special object (such as a frequently alarming pod) in a specific monitoring dimension.

    Directions

    This document describes how to automatically update a monitored object in the dashboard, automatically update an alarm object, and block a frequently triggered alarm object by taking the "TKE monitoring - pod" dimension as an example.

    Automatically updating monitored object in dashboard

    1. Log in to the Cloud Monitor console.
    2. Select Dashboard > Dashboard List > Create Dashboard > Create Chart.
    3. Configure the monitoring chart as detailed below:
      • Monitoring Type: select Cloud Product Monitoring here.
      • Metric: select "TKE (New) - pod" as the service and "CPU utilization (%)" as the metric here.
      • Filter: you can filter the objects to be bound to the chart by dimension (such as region, cluster, namespace, and workload).
        • Region: select the region of the monitored object.
        • Cluster: select the cluster of the monitored object.
      • Filter: you need to configure two filters, namely, the namespace and workload balance type, to monitor all pods under the specified workload and to automatically update the monitored object in the dashboard when pods are created/updated frequently, as shown below:
    4. After completing the configuration, click Save in the top-right corner of the page to save the chart.

    Automatically updating alarm object

    1. Log in to the Cloud Monitor console.
    2. Select Alarm Configuration > Alarm Policy > Create to enter the alarm policy creation page.
    3. Select "TKE (New) - pod" as the policy type and configure the alarm object as detailed below:
      • Region: select the region of the monitored object.
      • Cluster: select the cluster of the monitored object.
      • Filter: you need to configure two filters, namely, the namespace and workload balance type, to monitor all pods under the specified workload and to automatically update the alarm object when pods are created/updated frequently, as shown below:

        Note:

        For more information on how to configure an alarm, please see Creating Alarm Policy.

    Blocking frequently triggered alarm object

    When a pod frequently triggers an alarm, you can block some or all alarm objects under the node as instructed below.

    As shown below, you can block some pod alarms by configuring the "!=" operator for the pod name.