How to Set Ondat Container Resource Requests and Limits

Overview

With Ondat, cluster administrators can configure each individual Ondat container with specific requests and limits for CPU and memory resources.

Procedure

Edit the containerResources field in your storageoscluster spec to reflect your desired resources for the chosen containers. The below snippet is an example showing resource settings for the Ondat API-Manager and Ondat Scheduler.

spec:
  containerResources:
    apiManagerContainer:
      limits:
        cpu: 110m
        memory: 200Mi
      requests:
        cpu: 20m
        memory: 600Mi
    kubeSchedulerContainer:
      limits:
        cpu: 110m
        memory: 200Mi
      requests:
        cpu: 20m
        memory: 600Mi

If the cluster administrator performs this update successfully on a running cluster, the operator will restart the specified container(s) with the desired resource configurations.

Note: If the updated requests/limiits force a change in pod QoS (eg best-effort to guaranteed), then the affected deployment/daemonset will need to be manually deleted after the update. Once deleted, the operator will recreate the object with the correct resources and subsequent QoS.