Comment 0 for bug 1303890

The following exception was originally observed against the old rpc code, but the same problem exists in oslo.messaging.

 Traceback (most recent call last):
   File "/usr/lib/python2.6/site-packages/nova/openstack/common/excutils.py", line 78, in inner_func
     return infunc(*args, **kwargs)
   File "/usr/lib/python2.6/site-packages/nova/openstack/common/rpc/impl_qpid.py", line 698, in _consumer_thread
     self.consume()
   File "/usr/lib/python2.6/site-packages/nova/openstack/common/rpc/impl_qpid.py", line 689, in consume
     it.next()
   File "/usr/lib/python2.6/site-packages/nova/openstack/common/rpc/impl_qpid.py", line 606, in iterconsume
     yield self.ensure(_error_callback, _consume)
   File "/usr/lib/python2.6/site-packages/nova/openstack/common/rpc/impl_qpid.py", line 540, in ensure
     return method(*args, **kwargs)
   File "/usr/lib/python2.6/site-packages/nova/openstack/common/rpc/impl_qpid.py", line 597, in _consume
     nxt_receiver = self.session.next_receiver(timeout=timeout)
   File "<string>", line 6, in next_receiver
   File "/usr/lib/python2.6/site-packages/qpid/messaging/endpoints.py", line 665, in next_receiver
     if self._ecwait(lambda: self.incoming, timeout):
   File "/usr/lib/python2.6/site-packages/qpid/messaging/endpoints.py", line 50, in _ecwait
     result = self._ewait(lambda: self.closed or predicate(), timeout)
   File "/usr/lib/python2.6/site-packages/qpid/messaging/endpoints.py", line 571, in _ewait
     result = self.connection._ewait(lambda: self.error or predicate(), timeout)
   File "/usr/lib/python2.6/site-packages/qpid/messaging/endpoints.py", line 214, in _ewait
     self.check_error()
   File "/usr/lib/python2.6/site-packages/qpid/messaging/endpoints.py", line 207, in check_error
     raise self.error
 InternalError: Traceback (most recent call last):
   File "/usr/lib/python2.6/site-packages/qpid/messaging/driver.py", line 667, in write
     self._op_dec.write(*self._seg_dec.read())
   File "/usr/lib/python2.6/site-packages/qpid/framing.py", line 269, in write
     if self.op.headers is None:
 AttributeError: 'NoneType' object has no attribute 'headers'

It's possible for something to put the qpid client into a bad state. In particular, I have observed a case that will cause session.next_receiver() to immediately raise an AttributeError. This exception makes it all the way out. If the eventlet executor is used, the forever_retry_uncaught_exceptions() decorator will get hit. It will go back into this code and get the same error, stuck in an infinite loop of retrying.

The connection needs to be reset in this case to recover.