Uploaded image for project: 'ovsdb'
  1. ovsdb
  2. OVSDB-301

Various forms of VM restart cause floatingIp flows to be removed but then not replaced

XMLWordPrintable

    • Icon: Bug Bug
    • Resolution: Done
    • None
    • unspecified
    • openstack.net-virt
    • None
    • Operating System: All
      Platform: All

    • 5353

      The following tempest tests fail:
      tempest.scenario.test_network_advanced_server_ops.TestNetworkAdvancedServerOps.test_server_connectivity_resize
      tempest.scenario.test_network_advanced_server_ops.TestNetworkAdvancedServerOps.test_server_connectivity_stop_start
      tempest.scenario.test_network_advanced_server_ops.TestNetworkAdvancedServerOps.test_server_connectivity_suspend_resume
      tempest.scenario.test_network_advanced_server_ops.TestNetworkAdvancedServerOps.test_server_connectivity_rebuild
      tempest.scenario.test_network_advanced_server_ops.TestNetworkAdvancedServerOps.test_server_connectivity_reboot

      What these tests do is set up a VM, network, router, and a floating IP, reboot, rebuild, stop-start, etc., the VM and then test that the floating IP is still working.

      A SouthboundEvent is generated when the VM goes down which triggers the flow in table:30 to be removed. However, the corresponding SouthBoundEvents for when the port comes back up do not trigger the reconfiguration of the flow.

      This fix:
      https://git.opendaylight.org/gerrit/#/c/34394/
      is where the code to clean up the floating_ips when the port goes down was fixed.

            oshvartz@redhat.com Oded Shvartz
            jhershbe Josh Hershberg
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

              Created:
              Updated:
              Resolved: