rabbit_db: Run peer discovery from generic code

... instead of `rabbit_mnesia`.

[Why]
We need it for both Mnesia and Khepri. So instead of calling it in
`rabbit_khepri` too, let's manage this from `rabbit_db` directly.

deps/rabbit/src/rabbit_db.erl

@@ -55,6 +55,18 @@ init() ->
 
     pre_init(IsVirgin),
 
+    case IsVirgin of
+        true ->
+            %% At this point, the database backend could change if the node
+            %% joins a cluster and that cluster uses a different database.
+            ?LOG_INFO(
+               "DB: virgin node -> run peer discovery",
+               #{domain => ?RMQLOG_DOMAIN_DB}),
+            rabbit_peer_discovery:sync_desired_cluster();
+        false ->
+            ok
+    end,
+
     Ret = case rabbit_khepri:is_enabled() of
               true  -> init_using_khepri();
               false -> init_using_mnesia()

deps/rabbit/src/rabbit_mnesia.erl

@@ -114,17 +114,6 @@
 init() ->
     ensure_mnesia_running(),
     ensure_mnesia_dir(),
-    %% If this node is virgin, we call peer discovery to see if this node
-    %% should start as a standalone node or if it should join a cluster.
-    case is_virgin_node() of
-        true  ->
-            rabbit_log:info("Node database directory at ~ts is empty. "
-                            "Assuming we need to join an existing cluster or initialise from scratch...",
-                            [dir()]),
-            rabbit_peer_discovery:sync_desired_cluster();
-        false ->
-            ok
-    end,
     %% Peer discovery may have been a no-op if it decided that all other nodes
     %% should join this one. Therefore, we need to look at if this node is
     %% still virgin and finish our init of Mnesia accordingly. In particular,