Queue Manager fails in podman containerised environments

We hit this when trying to use queue manager in Kolla-Ansible. An invocation of nova-manage fails with the following traceback:

root@primary:~# podman exec nova_conductor nova-manage cell_v2 discover_hosts --by-service --cell_uuid c321ca04-f6e5-44dc-b257-b9ebac11a5eb
Traceback (most recent call last):
  File "/var/lib/kolla/venv/bin/nova-manage", line 8, in <module>
    sys.exit(main())
             ^^^^^^
  File "/var/lib/kolla/venv/lib/python3.12/site-packages/nova/cmd/manage.py", line 3815, in main
    config.parse_args(sys.argv)
  File "/var/lib/kolla/venv/lib/python3.12/site-packages/nova/config.py", line 101, in parse_args
    rpc.init(CONF)
  File "/var/lib/kolla/venv/lib/python3.12/site-packages/nova/rpc.py", line 68, in init
    TRANSPORT = create_transport(get_transport_url())
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/lib/kolla/venv/lib/python3.12/site-packages/nova/rpc.py", line 255, in create_transport
    return messaging.get_rpc_transport(CONF,
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/lib/kolla/venv/lib/python3.12/site-packages/oslo_messaging/rpc/transport.py", line 50, in get_rpc_transport
    return msg_transport._get_transport(
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/lib/kolla/venv/lib/python3.12/site-packages/oslo_messaging/transport.py", line 205, in _get_transport
    mgr = driver.DriverManager('oslo.messaging.drivers',
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/lib/kolla/venv/lib/python3.12/site-packages/stevedore/driver.py", line 54, in __init__
    super().__init__(
  File "/var/lib/kolla/venv/lib/python3.12/site-packages/stevedore/named.py", line 78, in __init__
    extensions = self._load_plugins(invoke_on_load,
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/lib/kolla/venv/lib/python3.12/site-packages/stevedore/extension.py", line 218, in _load_plugins
    self._on_load_failure_callback(self, ep, err)
  File "/var/lib/kolla/venv/lib/python3.12/site-packages/stevedore/extension.py", line 206, in _load_plugins
    ext = self._load_one_plugin(ep,
          ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/lib/kolla/venv/lib/python3.12/site-packages/stevedore/named.py", line 156, in _load_one_plugin
    return super()._load_one_plugin(
           ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/lib/kolla/venv/lib/python3.12/site-packages/stevedore/extension.py", line 242, in _load_one_plugin
    obj = plugin(*invoke_args, **invoke_kwds)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/lib/kolla/venv/lib/python3.12/site-packages/oslo_messaging/_drivers/impl_rabbit.py", line 1787, in __init__
    super().__init__(
  File "/var/lib/kolla/venv/lib/python3.12/site-packages/oslo_messaging/_drivers/amqpdriver.py", line 691, in __init__
    self._q_manager = QManager(
                      ^^^^^^^^^
  File "/var/lib/kolla/venv/lib/python3.12/site-packages/oslo_messaging/_drivers/amqpdriver.py", line 73, in __init__
    with open(f'/proc/{self.pg}/stat') as f:
         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
FileNotFoundError: [Errno 2] No such file or directory: '/proc/0/stat'

The issue here is that the group process id is zero when podman directly calls this command. /proc/0 doesn't exist, so the command fails.

podman exec nova_conductor ps xao pid,ppid,pgid,sid,comm
    PID PPID PGID SID COMMAND
      1 0 1 1 dumb-init
      2 1 1 1 nova-conductor
   2389 0 0 0 ps

As an aside, this doesn't affect docker as the process group id is not 0:

docker exec nova_conductor ps xao pid,ppid,pgid,sid,comm
    PID PPID PGID SID COMMAND
      1 0 1 1 dumb-init
      6 1 1 1 nova-conductor
     39 6 1 1 nova-conductor
     40 6 1 1 nova-conductor
   2418 0 2418 2418 ps