From 7b8fd6370c8b04c3836ee4f9d06eaa90c7be5197 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Herv=C3=A9=20Beraud?= Date: Fri, 3 May 2019 00:55:56 +0200 Subject: [PATCH] Fix switch connection destination when a rabbitmq cluster node disappear In a clustered rabbitmq when a node disappears, we get a ConnectionRefusedError because the socket get disconnected. The socket access yields a OSError because the heartbeat tries to reach an unreachable host (No route to host). Catch these exceptions to ensure that we call ensure_connection for switching the connection destination. POC is available at github.com:4383/rabbitmq-oslo_messging-error-poc Example: $ git clone git@github.com:4383/rabbitmq-oslo_messging-error-poc $ cd rabbitmq-oslo_messging-error-poc $ python -m virtualenv . $ source bin/activate $ pip install -r requirements.txt $ sudo podman run -d --hostname my-rabbit --name rabbit rabbitmq:3 $ python poc.py $(sudo podman inspect rabbit | niet '.[0].NetworkSettings.IPAddress') And in parallele in an another shell|tmux $ podman stop rabbit $ # observe the output of the poc.py script we now call ensure_connection Now you can observe some output relative to the connection who is modified and not catched before these changes. Related to: https://bugzilla.redhat.com/show_bug.cgi?id=1665399 Closes-Bug: #1828841 Change-Id: I9dc1644cac0e39eb11bf05f57bde77dcf6d42ed3 (cherry picked from commit 9d8b1430e5c081b081c0e3c0b5f12f744dc7809d) --- oslo_messaging/_drivers/impl_rabbit.py | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) diff --git a/oslo_messaging/_drivers/impl_rabbit.py b/oslo_messaging/_drivers/impl_rabbit.py index 5e07593aa..68f354312 100644 --- a/oslo_messaging/_drivers/impl_rabbit.py +++ b/oslo_messaging/_drivers/impl_rabbit.py @@ -897,6 +897,14 @@ class Connection(object): def _heartbeat_thread_job(self): """Thread that maintains inactive connections """ + # NOTE(hberaud): Python2 doesn't have ConnectionRefusedError + # defined so to switch connections destination on failure + # with python2 and python3 we need to wrapp adapt connection refused + try: + ConnectRefuseError = ConnectionRefusedError + except NameError: + ConnectRefuseError = socket.error + while not self._heartbeat_exit_event.is_set(): with self._connection_lock.for_heartbeat(): @@ -913,7 +921,17 @@ class Connection(object): self.connection.drain_events(timeout=0.001) except socket.timeout: pass + # NOTE(hberaud): In a clustered rabbitmq when + # a node disappears, we get a ConnectionRefusedError + # because the socket get disconnected. + # The socket access yields a OSError because the heartbeat + # tries to reach an unreachable host (No route to host). + # Catch these exceptions to ensure that we call + # ensure_connection for switching the + # connection destination. except (socket.timeout, + ConnectRefuseError, + OSError, kombu.exceptions.OperationalError) as exc: LOG.info(_LI("A recoverable connection/channel error " "occurred, trying to reconnect: %s"), exc)