Merged bug25474 and bug25486

1 files changed, 32 insertions, 5 deletions

diff --git a/src/rabbit_node_monitor.erl b/src/rabbit_node_monitor.erl
index 98e26a6a..de53b7f0 100644
--- a/src/rabbit_node_monitor.erl
+++ b/src/rabbit_node_monitor.erl
@@ -249,7 +249,8 @@ handle_info({'DOWN', _MRef, process, {rabbit, Node}, _Reason},
     write_cluster_status({AllNodes, DiscNodes, del_node(Node, RunningNodes)}),
     ok = handle_dead_rabbit(Node),
     [P ! {node_down, Node} || P <- pmon:monitored(Subscribers)],
-    {noreply, State#state{monitors = pmon:erase({rabbit, Node}, Monitors)}};
+    {noreply, handle_dead_rabbit_state(
+                State#state{monitors = pmon:erase({rabbit, Node}, Monitors)})};
 
 handle_info({'DOWN', _MRef, process, Pid, _Reason},
             State = #state{subscribers = Subscribers}) ->
@@ -257,10 +258,19 @@ handle_info({'DOWN', _MRef, process, Pid, _Reason},
 
 handle_info({mnesia_system_event,
              {inconsistent_database, running_partitioned_network, Node}},
-            State = #state{partitions = Partitions}) ->
+            State = #state{partitions = Partitions,
+                           monitors   = Monitors}) ->
+    %% We will not get a node_up from this node - yet we should treat it as
+    %% up (mostly).
+    State1 = case pmon:is_monitored({rabbit, Node}, Monitors) of
+                 true  -> State;
+                 false -> State#state{
+                            monitors = pmon:monitor({rabbit, Node}, Monitors)}
+             end,
+    ok = handle_live_rabbit(Node),
     Partitions1 = ordsets:to_list(
                     ordsets:add_element(Node, ordsets:from_list(Partitions))),
-    {noreply, State#state{partitions = Partitions1}};
+    {noreply, State1#state{partitions = Partitions1}};
 
 handle_info(_Info, State) ->
     {noreply, State}.
@@ -299,9 +309,14 @@ handle_dead_rabbit(Node) ->
     ok.
 
 majority() ->
+    length(alive_nodes()) / length(rabbit_mnesia:cluster_nodes(all)) > 0.5.
+
+%% mnesia:system_info(db_nodes) (and hence
+%% rabbit_mnesia:cluster_nodes(running)) does not give reliable results
+%% when partitioned.
+alive_nodes() ->
     Nodes = rabbit_mnesia:cluster_nodes(all),
-    Alive = [N || N <- Nodes, pong =:= net_adm:ping(N)],
-    length(Alive) / length(Nodes) > 0.5.
+    [N || N <- Nodes, pong =:= net_adm:ping(N)].
 
 await_cluster_recovery() ->
     rabbit_log:warning("Cluster minority status detected - awaiting recovery~n",
@@ -325,6 +340,18 @@ wait_for_cluster_recovery(Nodes) ->
                  wait_for_cluster_recovery(Nodes)
     end.
 
+handle_dead_rabbit_state(State = #state{partitions = Partitions}) ->
+    %% If we have been partitioned, and we are now in the only remaining
+    %% partition, we no longer care about partitions - forget them. Note
+    %% that we do not attempt to deal with individual (other) partitions
+    %% going away. It's only safe to forget anything about partitions when
+    %% there are no partitions.
+    Partitions1 = case Partitions -- (Partitions -- alive_nodes()) of
+                      [] -> [];
+                      _  -> Partitions
+                  end,
+    State#state{partitions = Partitions1}.
+
 handle_live_rabbit(Node) ->
     ok = rabbit_alarm:on_node_up(Node),
     ok = rabbit_mnesia:on_node_up(Node).