Nodes Management

    Cluster nodes are only accessible to cluster administrators. Administrators can find cluster nodes page by Cluster Adminstration -> Nodes -> Cluster Nodes . Some node metrics are very important to clusters, its administrators’ responsibilities to watch over these numbers to make sure nodes are available.

    • CPU : Node current cpu usage, these values are real-time numbers.
    • Memory : Current memory usage, same with CPU stats, are also real-time numbers.
    • Allocated Memory : Calculated from summing up all pods memory requests on this node. Same with Allocated CPU.

    Nodes Management

    Node Detail

    • Cordon/Uncordon : Marking a node as unschedulable are very useful during a node reboot or othere maintenance. The kubernetes scheduler will not schedule new pods to this node if it’s been marked unschedulable, but does not affect existing workloads already on this Node. In KubeSphere, you mark a node as unschedulable by click button Cordon in node detail page, node will be schedulable again if click the button again.

    Assign pods to nodes

    • Tainits : Taints allow node to repel a set of pods, . You can add or remove node taints in node detail page, but be careful, taints can cause unexpected behavior may lead to service unavailable.