documentation.suse.com / SUSE Enterprise Storage 7 Documentation / Deploying and Administering SUSE Enterprise Storage with Rook / Administrating Ceph on SUSE CaaS Platform / Ceph cluster administration
Applies to SUSE Enterprise Storage 7

3 Ceph cluster administration

  

This chapter introduces tasks that are performed on the whole cluster.

3.1 Shutting down and restarting the cluster

  

To shut down the whole Ceph cluster for planned maintenance tasks, follow these steps:

  1. Stop all clients that are using the cluster.

  2. Verify that the cluster is in a healthy state. Use the following commands: 

    cephuser@adm > ceph status
cephuser@adm > ceph health

  3. Set the following OSD flags: 

        cephuser@adm > ceph osd set noout
    cephuser@adm > ceph osd set nobackfill
    cephuser@adm > ceph osd set norecover

  4. Shutdown service nodes one by one (non-storage workers).

  5. Shutdown Ceph Monitor nodes one by one (masters by default).

  6. Shutdown Admin Node (masters).

After you finish the maintenance, you can start the cluster again by running the above procedure in reverse order.