Critical directories and services
---------------------------------
 - etcd database [ once ]
    * There is etcd2 and etcd3 APIs. OpenShift 3.5+ uses etcd3, but documentation
    still describes etcd2-style backup. etcd3 is backward compatible with etcd2,
    and we can run etcd2 backup as well. Now the question if we need to backup
    both ways (OpenShift 3.5 is definitively has etcd3 data) or just etcd3 
    considering it is a bug in documentation.
    * etcd3
        etcdctl3 --endpoints="192.168.213.1:2379" snapshot save snapshot.db
    * etcd2
        etcdctl backup --data-dir /var/lib/etcd/ --backup-dir .
        cp "$ETCD_DATA_DIR"/member/snap/db member/snap/db

 - heketi topology [ once ]
    heketi-cli -s  http://heketi-storage.glusterfs.svc.cluster.local:8080 --user admin --secret "$(oc get secret heketi-storage-admin-secret -n glusterfs -o jsonpath='{.data.key}' | base64 -d)" topology info --json 

 - Gluster volume information [ storage nodes ]
    * /var/lib/glusterd/glusterd.info
    * /var/lib/glusterd/peers
    * /var/lib/glusterd/glustershd                      - not mentioned in docs

  - etc [ all nodes ]
    * /etc/origin/                                      - Only *.key *.crt from /etc/origin/master in docs
    * /etc/etcd                                         - Not mentioned
    * /etc/docker                                       - Only certs.d