Description of problem:
rabbitmq has threads that are stuck even after restarting it after the /var/lib/rabbitmq/mnesia/ folder was emptied.

Version-Release number of selected component (if applicable):
rabbitmq-server-3.6.3-6.el7ost.noarch

How reproducible:
All the time on this environment.

Steps to Reproduce:
1. stop rabbitmq
2. empty /var/lib/rabbitmq/mnesia/ folder
3. start rabbitmq
4. run: rabbitmqctl eval 'rabbit_diagnostics:maybe_stuck().'

Actual results:
[root@controller1 ~]# rabbitmqctl eval 'rabbit_diagnostics:maybe_stuck().'
2019-08-16 18:38:33 There are 11728 processes.
2019-08-16 18:38:33 Investigated 4 processes this round, 5000ms to go.
2019-08-16 18:38:34 Investigated 4 processes this round, 4500ms to go.
2019-08-16 18:38:34 Investigated 4 processes this round, 4000ms to go.
2019-08-16 18:38:35 Investigated 4 processes this round, 3500ms to go.
2019-08-16 18:38:35 Investigated 4 processes this round, 3000ms to go.
2019-08-16 18:38:36 Investigated 4 processes this round, 2500ms to go.
2019-08-16 18:38:36 Investigated 4 processes this round, 2000ms to go.
2019-08-16 18:38:37 Investigated 4 processes this round, 1500ms to go.
2019-08-16 18:38:37 Investigated 4 processes this round, 1000ms to go.
2019-08-16 18:38:38 Investigated 4 processes this round, 500ms to go.
2019-08-16 18:38:38 Found 4 suspicious processes.
2019-08-16 18:38:38 [{pid,<10374.16909.97>},
                     {registered_name,[]},
                     {current_stacktrace,
                         [{rabbit_control_misc,collect_monitors,1,
                              [{file,"src/rabbit_control_misc.erl"},
                               {line,124}]},
                          {rpc,'-handle_call_call/6-fun-0-',5,
                              [{file,"rpc.erl"},{line,206}]}]},
                     {initial_call,{erlang,apply,2}},
                     {message_queue_len,3},
                     {links,[]},
                     {monitors,[]},
                     {monitored_by,[<10374.14.0>]},
                     {heap_size,233}]
2019-08-16 18:38:38 [{pid,<10374.4211.98>},
                     {registered_name,[]},
                     {current_stacktrace,
                         [{rabbit_control_misc,collect_monitors,1,
                              [{file,"src/rabbit_control_misc.erl"},
                               {line,124}]},
                          {rpc,'-handle_call_call/6-fun-0-',5,
                              [{file,"rpc.erl"},{line,206}]}]},
                     {initial_call,{erlang,apply,2}},
                     {message_queue_len,3},
                     {links,[]},
                     {monitors,[]},
                     {monitored_by,[<10374.14.0>]},
                     {heap_size,233}]
2019-08-16 18:38:38 [{pid,<10374.9131.98>},
                     {registered_name,[]},
                     {current_stacktrace,
                         [{rabbit_control_misc,collect_monitors,1,
                              [{file,"src/rabbit_control_misc.erl"},
                               {line,124}]},
                          {rpc,'-handle_call_call/6-fun-0-',5,
                              [{file,"rpc.erl"},{line,206}]}]},
                     {initial_call,{erlang,apply,2}},
                     {message_queue_len,3},
                     {links,[]},
                     {monitors,[]},
                     {monitored_by,[<10374.14.0>]},
                     {heap_size,233}]
[...]

Expected results:
no threads stuck

Additional info:
this affects all openstack services that depends on rabbitmq for inter-communication