Hey,

i am getting this error after an upgrade to pike.

    Traceback (most recent call last):
2019-06-25 10:42:10.965 15041 ERROR nova.compute.manager    File "/usr/lib/python2.7/dist-packages/nova/compute/manager.py", line 6150, in _heal_instance_info_cache
2019-06-25 10:42:10.965 15041 ERROR nova.compute.manager      self.network_api.get_instance_nw_info(context, instance)
2019-06-25 10:42:10.965 15041 ERROR nova.compute.manager    File "/usr/lib/python2.7/dist-packages/nova/network/base_api.py", line 255, in get_instance_nw_info
2019-06-25 10:42:10.965 15041 ERROR nova.compute.manager      result = self._get_instance_nw_info(context, instance, **kwargs)
2019-06-25 10:42:10.965 15041 ERROR nova.compute.manager    File "/usr/lib/python2.7/dist-packages/nova/network/neutronv2/api.py", line 1322, in _get_instance_nw_info
2019-06-25 10:42:10.965 15041 ERROR nova.compute.manager      preexisting_port_ids)
2019-06-25 10:42:10.965 15041 ERROR nova.compute.manager    File "/usr/lib/python2.7/dist-packages/nova/network/neutronv2/api.py", line 2267, in _build_network_info_model
2019-06-25 10:42:10.965 15041 ERROR nova.compute.manager      client = get_client(context, admin=True)
2019-06-25 10:42:10.965 15041 ERROR nova.compute.manager    File "/usr/lib/python2.7/dist-packages/nova/network/neutronv2/api.py", line 167, in get_client
2019-06-25 10:42:10.965 15041 ERROR nova.compute.manager      global_request_id=context.global_id),
2019-06-25 10:42:10.965 15041 ERROR nova.compute.manager  AttributeError: 'RequestContext' object has no attribute 'global_id'

This shows up on the hypervisor. i cannot figure out why. Since both hypervisor and cotnroller have the same version and i resetet the rabbitmq to make sure there is no wrong data stored in the rabbitmq.

Any ideas where that comes from?

Happens if i want to start VMs after the upgrade and i cannot create new VMs

thanks