Config - Monitoring Clusters

Rest API Config endpoint for monitoring clusters

Object type: monitoringcluster

Request URL: /rest/config/monitoringcluster

Example GET:

{
   "object" : {
      "roles" : [
         {
            "ref" : "/rest/config/role/12",
            "name" : "View some, change some"
         }
      ],
      "activated" : "1",
      "monitors" : [
         {
            "ref" : "/rest/config/host/10",
            "name" : "cisco3"
         },
         {
            "ref" : "/rest/config/host/4",
            "name" : "monitored_remotely"
         }
      ],
      "name" : "ClusterA",
      "nodes" : [
         {
            "host" : {
               "ip" : "192.168.10.20",
               "ref" : "/rest/config/host/5",
               "name" : "collector1"
            },
            "slave_port" : "22"   // Unused in 6.0
         }
      ],
      "id" : "2",
      "uncommitted" : "1"
   }
}

If id=1, this is the primary monitoring cluster.

DELETEs are blocked if the monitoring cluster is the primary, or if there are any hosts still monitored by this monitoring cluster.

Additional parameters:

  • order - can be num_hosts or num_nodes to order the list of monitoring clusters by those columns
  • include_delete_info - if set, will return extra fields indicating if a monitoring cluster is not deletable because it is the primary, it is used as a netflow collector, it still has hosts associated to it, or if the monitoring cluster is license locked
  • include_cluster_details - if set to 1, then will include two additional attributes in each monitoring cluster’s data - see below

The additional details included in the output when include_cluster_details is used are:

"status": "OK",    // Other values: OFFLINE or DEGRADED
"alarms": [        // Contains an array of strings of alarms about this cluster
                   // Note: Strings will be localised based on language
  "Component 'opsview-scheduler' is OFFLINE", 
  "No response in time"
]

There is an activated_calculated column that will be included if order=activated_calculated is set. This can be one of 4 values (note that value 3 is not ordered correctly - this is a known limitation):

  • 0 = Not activated
  • 1 = Activated cluster
  • 2 = Always activated (primary)
  • 3 = Disabled due to licensing

When PUT/POSTing to monitoringclusters, monitors and roles is unsupported. The nodes parameter should be of the form:

nodes: [ { id: 13 }, { id: 20 } ]

where the ids are the host ids of the collectors. For the primary monitoring cluster, only the first node is used and the remainder silently ignored.