Remote metadata server internal server error

[Peter Sabaini]

Getting intermittent issues getting connectivity when spawning instances. The instance would spawn ok but never leave 'pending' state from Jujus POV.

On the console I can see it doesn't get a hostname set, and on the compute host I see these tracebacks, from neutron-ovn-metadata-agent.log:

2023-03-08 13:46:16.025 3879746 INFO eventlet.wsgi.server [-] 10.5.0.185,<local> "GET /openstack HTTP/1.1" status: 500 len: 0 time: 60.1910186
2023-03-08 13:46:28.683 3879746 WARNING neutron.agent.ovn.metadata.server [-] Remote metadata server experienced an internal server error.
2023-03-08 13:46:28.685 3879746 INFO eventlet.wsgi.server [-] Traceback (most recent call last):
  File "/usr/lib/python3/dist-packages/eventlet/wsgi.py", line 604, in handle_one_response
    write(b''.join(towrite))
  File "/usr/lib/python3/dist-packages/eventlet/wsgi.py", line 538, in write
    wfile.flush()
  File "/usr/lib/python3.8/socket.py", line 687, in write
    return self._sock.send(b)
  File "/usr/lib/python3/dist-packages/eventlet/greenio/base.py", line 399, in send
    return self._send_loop(self.fd.send, data, flags)
  File "/usr/lib/python3/dist-packages/eventlet/greenio/base.py", line 386, in _send_loop
    return send_method(data, *args)
BrokenPipeError: [Errno 32] Broken pipe

This is focal running:
neutron-ovn-metadata-agent 2:20.2.0-0ubuntu1~cloud0

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.