Skip to content

Allow a previously reset node to rejoin its original cluster #13643

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 6 commits into from
Apr 1, 2025
Merged

Allow a previously reset node to rejoin its original cluster #13643

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
dd49cbe
Mnesia: Ask to leave a cluster and retry to join if cluster already c…
SimonUnge Mar 27, 2025
9ba545c
Fix dialyzer issue.
SimonUnge Mar 31, 2025
e1f2865
Return the exception
SimonUnge Mar 31, 2025
cdeabe2
Dont handle the exception just let it out there
SimonUnge Mar 31, 2025
36eb6ca
Update spec, noconnection is also a possible error
SimonUnge Mar 31, 2025
e6bc6a4
Naming #13643
michaelklishin Apr 1, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 22 additions & 0 deletions deps/rabbit/src/rabbit_db_cluster.erl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -214,6 +214,28 @@ join(RemoteNode, NodeType)
end;
{ok, already_member} ->
{ok, already_member};
{error, {inconsistent_cluster, Msg}} = Error ->
case rabbit_khepri:is_enabled() of
true ->
Error;
false ->
%% rabbit_mnesia:can_join_cluster/1 notice inconsistent_cluster,
%% as RemoteNode thinks this node is already in the cluster.
%% Attempt to leave the RemoteNode cluster, the discovery cluster,
%% and simply retry the operation.
rabbit_log:info("Mnesia: node ~tp thinks it's clustered "
"with node ~tp, but ~tp disagrees. ~tp will ask "
"to leave the cluster and try again.",
[RemoteNode, node(), node(), node()]),
try
ok = rabbit_mnesia:leave_discover_cluster(RemoteNode),
join(RemoteNode, NodeType)
catch
_ ->
rabbit_log:error(Msg),
Error
end
end;
{error, _} = Error ->
Error
end.
Expand Down
23 changes: 13 additions & 10 deletions deps/rabbit/src/rabbit_mnesia.erl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -73,7 +73,7 @@
-export([node_info/0, remove_node_if_mnesia_running/1]).

%% Used internally in `rabbit_db_cluster'.
-export([members/0]).
-export([members/0, leave_discover_cluster/1]).

%% Used internally in `rabbit_khepri'.
-export([mnesia_and_msg_store_files/0]).
Expand Down Expand Up @@ -179,7 +179,6 @@ can_join_cluster(DiscoveryNode) ->
{ok, already_member};
false ->
Msg = format_inconsistent_cluster_message(DiscoveryNode, node()),
rabbit_log:error(Msg),
{error, {inconsistent_cluster, Msg}}
end
end.
Expand Down Expand Up @@ -923,15 +922,19 @@ remove_node_if_mnesia_running(Node) ->
end
end.

leave_cluster() ->
case rabbit_nodes:nodes_excl_me(cluster_nodes(all)) of
[] -> ok;
AllNodes -> case lists:any(fun leave_cluster/1, AllNodes) of
true -> ok;
false -> e(no_running_cluster_nodes)
end
end.
leave_discover_cluster(DiscoveryNode) ->
{ClusterNodes, _, _} = discover_cluster([DiscoveryNode]),
leave_cluster(rabbit_nodes:nodes_excl_me(ClusterNodes)).

leave_cluster() ->
leave_cluster(rabbit_nodes:nodes_excl_me(cluster_nodes(all))).
leave_cluster([]) ->
ok;
leave_cluster(Nodes) when is_list(Nodes) ->
case lists:any(fun leave_cluster/1, Nodes) of
true -> ok;
false -> e(no_running_cluster_nodes)
end;
leave_cluster(Node) ->
case rpc:call(Node,
rabbit_mnesia, remove_node_if_mnesia_running, [node()]) of
Expand Down
Loading