neutron-openvswitch-agent report error when batch delete instances
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
neutron |
Fix Released
|
Undecided
|
Zachary Ma |
Bug Description
I created some instance with same network and same security group. The security group's remote secrity group is set to itself.
Then I batch delete these instances.when the first instance's port is deleted,it will notify other neutron-ovs-agent so that these agent could update filter rules for other instance's port.
But variable "defer_
During the subsequent instance deletion,the process will be triggered,But the port within subsequent instance is deleted,So the agent will report error.
2017-11-01 09:30:29.695 9779 ERROR neutron.
summary: |
- neutron-ovs-agnet report error when batch delete instances + neutron-openvswitch-agnet report error when batch delete instances |
summary: |
- neutron-openvswitch-agnet report error when batch delete instances + neutron-openvswitch-agent report error when batch delete instances |
Changed in neutron: | |
status: | New → Confirmed |
tags: | added: neutron-proactive-backport-potential |
tags: |
added: ovs-fw removed: neutron-proactive-backport-potential |
in function as follow: create_ ofport( self, port): ovs_port( port_id)
def get_or_
"""Get ofport specified by port['device'], checking and reflecting
ofport changes.
If ofport is nonexistent, create and return one.
"""
port_id = port['device']
ovs_port = self.get_
libvirt will delete tap when removed vms .then, ovs agent check ovs_port by "self.get_ovs_port" , and the exception will be raised.