Chapter 8. Investigating HA Ceph Nodes

When deployed with Ceph storage, Red Hat OpenStack Platform uses ceph-mon as a monitor daemon for the Ceph cluster. The director deploys this daemon on all controller nodes.

To check whether the Ceph Monitoring service is running, use:

$ sudo service ceph status
=== mon.overcloud-controller-0 ===
mon.overcloud-controller-0: running {"version":"0.94.1"}

On the controllers, as well as on the Ceph Nodes, you can see how Ceph is configured by viewing the /etc/ceph/ceph.conf file. For example:

[global]
osd_pool_default_pgp_num = 128
osd_pool_default_min_size = 1
auth_service_required = cephx
mon_initial_members = overcloud-controller-0,overcloud-controller-1,overcloud-controller-2
fsid = 8c835acc-6838-11e5-bb96-2cc260178a92
cluster_network = 172.19.0.11/24
auth_supported = cephx
auth_cluster_required = cephx
mon_host = 172.18.0.17,172.18.0.15,172.18.0.16
auth_client_required = cephx
osd_pool_default_size = 3
osd_pool_default_pg_num = 128
public_network = 172.18.0.17/24

Here, all three controller nodes (overcloud-controller-0, -1, and -2) are set to monitor the Ceph cluster (mon_initial_members). The 172.19.0.11/24 network (VLAN 203) is used as the Storage Management Network and provides a communications path between the controller and Ceph Storage Nodes.

The three Ceph Storage Nodes are on a separate network. As you can see, the IP addresses for those three nodes are on the Storage Network (VLAN 202) and are defined as 172.18.0.15, 172.18.0.16, and 172.18.0.17.

To check the current status of a Ceph node, log into that node and run the following command:

# ceph -s
    cluster 8c835acc-6838-11e5-bb96-2cc260178a92
     health HEALTH_OK
     monmap e1: 3 mons at {overcloud-controller-0=172.18.0.17:6789/0,overcloud-controller-1=172.18.0.15:6789/0,overcloud-controller-2=172.18.0.16:6789/0}
            election epoch 152, quorum 0,1,2 overcloud-controller-1,overcloud-controller-2,overcloud-controller-0
     osdmap e543: 6 osds: 6 up, 6 in
      pgmap v1736: 256 pgs, 4 pools, 0 bytes data, 0 objects
            267 MB used, 119 GB / 119 GB avail
                 256 active+clean

From the ceph -s output, you can see that the health of the Ceph cluster is OK (HEALTH_OK). There are three Ceph monitor services running on the three overcloud-controller nodes. Also shown here are the IP addresses and ports each is listening on.

For more information about Red Hat Ceph, see the Red Hat Ceph product page.