monitoring: fix 'CephMonQuorumLost' alert

Only the 'Running' mons with result value of '1' should be counted. Signed-off-by: Arun Kumar Mohan <amohan@redhat.com>
rook · Nov 2, 2021 · af44e5c · af44e5c
1 parent 7375679
commit af44e5c
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/cluster/examples/kubernetes/ceph/monitoring/prometheus-ceph-v14-rules.yaml b/cluster/examples/kubernetes/ceph/monitoring/prometheus-ceph-v14-rules.yaml
@@ -90,7 +90,7 @@ spec:
         severity_level: critical
         storage_type: ceph
       expr: |
-        count(kube_pod_status_phase{pod=~"rook-ceph-mon-.*", phase=~"Running|running"}) by (namespace) < 2
+        count(kube_pod_status_phase{pod=~"rook-ceph-mon-.*", phase=~"Running|running"} == 1) by (namespace) < 2
       for: 5m
       labels:
         severity: critical