feat: cluster info available in metrics

apps/emqx_prometheus/src/emqx_prometheus.erl

@@ -124,8 +124,10 @@ collect_mf(_Registry, Callback) ->
     Metrics = emqx_metrics:all(),
     Stats = emqx_stats:getstats(),
     VMData = emqx_vm_data(),
+    ClusterData = emqx_cluster_data(),
     _ = [add_collect_family(Name, Stats, Callback, gauge) || Name <- emqx_stats()],
     _ = [add_collect_family(Name, VMData, Callback, gauge) || Name <- emqx_vm()],
+    _ = [add_collect_family(Name, ClusterData, Callback, gauge) || Name <- emqx_cluster()],
     _ = [add_collect_family(Name, Metrics, Callback, counter) || Name <- emqx_metrics_packets()],
     _ = [add_collect_family(Name, Metrics, Callback, counter) || Name <- emqx_metrics_messages()],
     _ = [add_collect_family(Name, Metrics, Callback, counter) || Name <- emqx_metrics_delivery()],
@@ -454,7 +456,13 @@ emqx_collect(emqx_vm_total_memory, VMData) ->
     gauge_metric(?C(total_memory, VMData));
 
 emqx_collect(emqx_vm_used_memory, VMData) ->
-    gauge_metric(?C(used_memory, VMData)).
+    gauge_metric(?C(used_memory, VMData));
+
+emqx_collect(emqx_cluster_nodes_running, ClusterData) ->
+    gauge_metric(?C(nodes_running, ClusterData));
+
+emqx_collect(emqx_cluster_nodes_stopped, ClusterData) ->
+    gauge_metric(?C(nodes_stopped, ClusterData)).
 
 %%--------------------------------------------------------------------
 %% Indicators
@@ -592,3 +600,13 @@ emqx_vm_data() ->
      {process_total_messages, 0}, %% XXX: Plan removed at v5.0
      {cpu_idle, Idle},
      {cpu_use, 100 - Idle}] ++ emqx_vm:mem_info().
+
+emqx_cluster() ->
+    [ emqx_cluster_nodes_running
+    , emqx_cluster_nodes_stopped
+    ].
+
+emqx_cluster_data() ->
+    #{running_nodes := Running, stopped_nodes := Stopped} = ekka_mnesia:cluster_info(),
+    [{nodes_running, length(Running)},
+     {nodes_stopped, length(Stopped)}].