[Rook] Disscusion about template changes for non-openstack Deployments

Comparing the three non-openstack deployments we already set up, I came up with a preliminary list of required changes:

  • specific placement label for mons and mgrs (MR: !499 (merged)) e.g.: rook_{mon,mgr}_scheduling_key
  • an adjustable number of mons and mgr operating in the ceph cluster e.g.: via rook_nomons and rook_nomgrs (MR: !477 (merged))
  • exclude the storageClassDeviceSets and volumeClaimTemplates (included in MR: !504 (merged))
  • have the ability to enable host networking (MR: !484 (merged))
  • set a specific for CSI_PLUGIN_NODE_AFFINITY label in the rook operator template disable csi-plugins if the rook cluster is not used for k8s workloads and do not create the associated storageclass (MR:!495 (merged))
  • enable or disable encryption for devices (via: encryptedDevice {true|false}) (MR: !497 (merged))
  • give a list or a dict with nodes and devices + configuration to use or set to use all nodes and devices in the storage section (MR: !504 (merged)) e.g.:
  storage: # cluster level storage configuration and selection
    useAllNodes: false
    useAllDevices: false
    nodes:
    - name: "mXXrY"
      devices: # specific devices to use for storage can be specified for each node
      - name: "sdZ"
      - name: "sdZ"
Edited by Tobias Fischer
To upload designs, you'll need to enable LFS and have an admin enable hashed storage. More information