CLI: s/metadata_store/rabbitmq_metadata/

Fixes error handling and a `clsutering_management_SUITE` transient
failure.

Author: dumbbell

deps/rabbit/test/clustering_management_SUITE.erl

@@ -614,13 +614,16 @@ forget_unavailable_node_in_minority(Config) ->
        30000),
 
     %% If Hare was the leader, it is able to forget one of the nodes. Change takes place as soon as it is written on the log. The other membership change will be rejected until the last change has consensus.
+    ct:pal("Initial Raft status: ~p", [RaftStatus]),
     case RaftStatus of
         leader ->
             ?assertMatch(ok, forget_cluster_node(Config, Hare, Rabbit)),
             not_permitted(forget_cluster_node(Config, Hare, Bunny));
         follower ->
             %% Follower might have been promoted before the second node goes down, check the status again
-            case get_raft_status(Config, Hare) of
+            RaftStatus1 = get_raft_status(Config, Hare),
+            ct:pal("Latest Raft status: ~p", [RaftStatus1]),
+            case RaftStatus1 of
                 leader ->
                     ?assertMatch(ok, forget_cluster_node(Config, Hare, Rabbit)),
                     not_permitted(forget_cluster_node(Config, Hare, Bunny));

deps/rabbitmq_cli/lib/rabbitmq/cli/default_output.ex

@@ -68,7 +68,7 @@ defmodule RabbitMQ.CLI.DefaultOutput do
   defp normalize_output({unknown, _} = input, _opts) when is_atom(unknown), do: {:error, input}
   defp normalize_output(result, _opts) when not is_atom(result), do: {:ok, result}
 
-  defp format_khepri_output({:error, {:timeout, {:metadata_store, _}}}, %{node: node_name}) do
+  defp format_khepri_output({:error, {:timeout, {:rabbitmq_metadata, _}}}, %{node: node_name}) do
     {:error, RabbitMQ.CLI.Core.ExitCodes.exit_tempfail(), khepri_timeout_error(node_name)}
   end