Appendix D. Manually installing Ceph Manager

Usually, the Ansible automation utility installs the Ceph Manager daemon (ceph-mgr) when you deploy the Red Hat Ceph Storage cluster. However, if you do not use Ansible to manage Red Hat Ceph Storage, you can install Ceph Manager manually. Red Hat recommends to colocate the Ceph Manager and Ceph Monitor daemons on a same node.


  • A working Red Hat Ceph Storage cluster
  • root or sudo access
  • The rhel-7-server-rhceph-3-mon-rpms repository enabled
  • Open ports 6800-7300 on the public network if firewall is used


Use the following commands on the node where ceph-mgr will be deployed and as the root user or with the sudo utility.

  1. Install the ceph-mgr package:

    [root@node1 ~]# yum install ceph-mgr
  2. Create the /var/lib/ceph/mgr/ceph-hostname/ directory:

    mkdir /var/lib/ceph/mgr/ceph-hostname

    Replace hostname with the host name of the node where the ceph-mgr daemon will be deployed, for example:

    [root@node1 ~]# mkdir /var/lib/ceph/mgr/ceph-node1
  3. In the newly created directory, create an authentication key for the ceph-mgr daemon:

    [root@node1 ~]# ceph auth get-or-create mgr.`hostname -s` mon 'allow profile mgr' osd 'allow *' mds 'allow *' -o /var/lib/ceph/mgr/ceph-node1/keyring
  4. Change the owner and group of the /var/lib/ceph/mgr/ directory to ceph:ceph:

    [root@node1 ~]# chown -R ceph:ceph /var/lib/ceph/mgr
  5. Enable the ceph-mgr target:

    [root@node1 ~]# systemctl enable
  6. Enable and start the ceph-mgr instance:

    systemctl enable ceph-mgr@hostname
    systemctl start ceph-mgr@hostname

    Replace hostname with the host name of the node where the ceph-mgr will be deployed, for example:

    [root@node1 ~]# systemctl enable ceph-mgr@node1
    [root@node1 ~]# systemctl start ceph-mgr@node1
  7. Verify that the ceph-mgr daemon started successfully:

    ceph -s

    The output will include a line similar to the following one under the services: section:

        mgr: node1(active)
  8. Install more ceph-mgr daemons to serve as standby daemons that become active if the current active daemon fails.