Why pacemaker restoring from cluster standby cause multiple OVS port to be tagged 4095 (dead) on HA OpenStack environment ?
Issue
- Intermittently , restoring pacemaker cluster from standby results in one or more OVS ports tagged 4095 (dead) .
# pcs cluster standby --all
# pcs cluster unstandby --all
# watch -n2 "ovs-vsctl show | grep -B1 tag"
- The dead port tagged 4095 can be recovered by restarting ovs agent on controller .
# pcs resource restart neutron-openvswitch-agent-clone
Environment
- Red Hat OpenStack 6
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.