Since end of 2014 IPv6 traffic causes host to reboot

[Imad Benbrahim]

We launch IPv6 traffic to a VM within an Openstack Host that has icehouse. Soon the host restarts. In /var/log/messages and /var/log/openvswitch/ovs-vswitchd.log we see multiple instances of this error:
      ovs-vswitchd: ovs|00025|odp_util(revalidator_20)|ERR|mask expected for non-Ethernet II frame

We tried the same test on multiple other machines that have slightly different versions of neutron. Our conclusion is that the bug must have been introduced with :
    openstack-neutron.noarch 2014.1.3-5.el6
    openstack-neutron-ml2.noarch 2014.1.3-5.el6
    openstack-neutron-openvswitch.noarch 2014.1.3-5.el6
    python-neutron.noarch 2014.1.3-5.el6

Machines with the above version (or later) will exhibit the bug.

Machines with the following version will work fine:
    openstack-neutron.noarch 2014.1.3-4.el6
    openstack-neutron-ml2.noarch 2014.1.3-4.el6
    openstack-neutron-openvswitch.noarch 2014.1.3-4.el6
    python-neutron.noarch 2014.1.3-4.el6

[Sean M. Collins (scollins)]

What version of Open vSwitch are you running? I haven't looked at your logs to see if it prints the version, but if the host is rebooting- that is most likely due to a kernel panic

[Imad Benbrahim (ibenbrahim)]

openvswitch version is identical on both machines: 2.1.2-1.el6.
The theory about neutron version 2014.1.3-4 working better may not hold according the latest tests. The problem seems to happen randomly in some machines that have icehouse on top of CentOS 6.5.

[Armando Migliaccio (armando-migliaccio)]

This bug is > 172 days without activity. We are unsetting assignee and milestone and setting status to Incomplete in order to allow its expiry in 60 days.

If the bug is still valid, then update the bug status.

[Launchpad Janitor (janitor)]

[Expired for neutron because there has been no activity for 60 days.]