Fix proper templating when zookeeper service down on some nodes

Change-Id: I922331e19bf2d6db2b61d0ea725c8ca0965e7e8b
Depends-On: I80b43221d39048cf2268b59e8b9cd70f2b693b78
diff --git a/opencontrail/files/grafana_dashboards/zookeeper_prometheus.json b/opencontrail/files/grafana_dashboards/zookeeper_prometheus.json
index 2fbcf39..8be6312 100644
--- a/opencontrail/files/grafana_dashboards/zookeeper_prometheus.json
+++ b/opencontrail/files/grafana_dashboards/zookeeper_prometheus.json
@@ -1083,7 +1083,7 @@
         "multi": true,
         "name": "host",
         "options": [],
-        "query": "label_values(zookeeper_znode_count,host)",
+        "query": "label_values(procstat_running{process_name=\"zookeeper-server\"},host)",
         "refresh": 1,
         "refresh_on_load": true,
         "regex": "",