Fix _node/$node/_system endpoint

The endpoint is using same handler as a backdoor interface,
which only works for the requests to a local $node.
For rpc calls to the remote nodes, the same as it's done
on rest of /_node endpoints, handle_node_req handler is getting
{badrpc,{'EXIT',normal}} response and fails with invalid_ejson exception.

The change splits stats gathering from stats responding,
allowing to use former on both /_node/$node/_system
and backend's /_system endpoints.
1 file changed
tree: 8a00d4bffb91bcef534e3e845876184412ef89ef
  1. include/
  2. priv/
  3. src/
  4. test/
  5. .travis.yml
  6. LICENSE