ceph: adding 'namespace' into the result of 'ceph_node_down' query

This will add 'namespace' field into the result of 'ceph_node_down'
query.

Signed-off-by: Arun Kumar Mohan <amohan@redhat.com>

cluster/examples/kubernetes/ceph/monitoring/prometheus-ceph-v14-rules.yaml

@@ -11,7 +11,7 @@ spec:
   - name: ceph.rules
     rules:
     - expr: |
-        kube_node_status_condition{condition="Ready",job="kube-state-metrics",status="true"} * on (node) group_right() max(label_replace(ceph_disk_occupation{job="rook-ceph-mgr"},"node","$1","exported_instance","(.*)")) by (node)
+        kube_node_status_condition{condition="Ready",job="kube-state-metrics",status="true"} * on (node) group_right() max(label_replace(ceph_disk_occupation{job="rook-ceph-mgr"},"node","$1","exported_instance","(.*)")) by (node, namespace)
       record: cluster:ceph_node_down:join_kube
     - expr: |
         avg(topk by (ceph_daemon) (1, label_replace(label_replace(ceph_disk_occupation{job="rook-ceph-mgr"}, "instance", "$1", "exported_instance", "(.*)"), "device", "$1", "device", "/dev/(.*)")) * on(instance, device) group_right(ceph_daemon) topk by (instance,device) (1,(irate(node_disk_read_time_seconds_total[1m]) + irate(node_disk_write_time_seconds_total[1m]) / (clamp_min(irate(node_disk_reads_completed_total[1m]), 1) + irate(node_disk_writes_completed_total[1m])))))