Skip to content

CephOsdRemoveRequest#

The CephOsdRemoveRequest object is used for Ceph Object Storage Daemon (OSD) lifecycle management (LCM). It allows the operator to manage the removal of OSDs from the cluster.

Spec#

The spec section describes what should be cleaned up:

  • nodes: A mapping of node names with specifications:
    • completeCleanUp: Boolean flag for full node cleanup (including removal from the CRUSH map).
    • dropFromClusterOnly: Remove the OSD from the cluster without cleaning devices.
    • cleanupStrayPartitionsOnly: Clean only stray partitions.
    • cleanupByOsdId: List of OSD IDs to remove.
    • cleanupByDevice: List of devices to clean (by name or path).
  • Additional fields include approve, keepOnFail, resolved, resumeFailed, and resumeFailedStep.

Example:

apiVersion: lcm.mirantis.com/v1alpha1
kind: CephOsdRemoveRequest
metadata:
  name: lcm-req
  namespace: ceph-lcm-mirantis
spec:
  nodes:
    node-a:
      completeCleanUp: true
    node-b:
      cleanupByDevice:
      - name: sdb
      - path: "/dev/disk/by-path/pci-0000:00:0x.0"
    node-c:
      cleanupByOsdId:
      - 5
      - 10

Status#

The status section provides: - phase: Current phase of the request (e.g., Pending, Processing, CompletedWithWarnings). - phaseInfo: Additional information about the phase. - removeInfo: Detailed mapping of what is scheduled for removal. - failedStepID: Identifier of any failed step. - messages: List of messages regarding request progress. - conditions: Historical record of phase transitions and status changes.

(A full example is provided in the document.)