Configuring CPU and memory resources for FDF-managed pods

You can define and adjust resource requests or limits for the pods managed by Fusion Data Foundation (FDF) to ensure optimal scheduling and performance of FDF workloads.

Configuring resources for operator pods

You can configure resources for operator pods, such as, ocs-operator, odf-operator, noobaa-operator, ocs-client-operator, and so on. You can edit the default resources in their Subscription as follows:
kind: Subscription
metadata:
  name: my-operator
spec:
  package: my-operator-package
  channel: alpha
  config:
    resources:
      requests:
        memory: <value>
        cpu: <value>
      limits:
        memory: <value>
        cpu: <value>

Configuring resources for pods deployed by ocs-operator

You can configure sub-resource pods deployed by ocs-operator, such as rook-ceph-mgr, rook-ceph-mon, ocs-metrics-exporter and so on. You can update the values for these pods by editing the resources field in the storage cluster as follows:
spec:
  resources:
    ocs-metrics-exporter:
      requests:
        cpu: <value>
        memory: <value>
      limits:
        cpu: <value>
        memory: <value>

Configuring CSI pods managed by ceph-csi-operator

You can configure CSI pods in the openshift-storage namespace that are managed by the ceph-csi-operator. You can edit the drivers.csi.ceph.io CR for each driver in the openshift-storage namespace and update the resource values for ControllerPlugin/NodePlugin resources based on the requirement as follows:
spec:
  controllerPlugin:
    resources:
      snapshotter:
        requests:
          cpu: <value>
          memory: <value>
        limits:
          cpu: <value>
          memory: <value>