VXLAN tunnel to the controller is not being created on a compute node after perform multiple vm autoevacuations

Solution Verified - Updated -

Issue

Launch a stack, and run vm autoevacuation test by rebooting one of the compute node. After the compute node boots up, since there are no vms on this compute, vxlan tunnels are vanished, then reboot another compute node to trigger the second vm autoevacuation.As a result, some vms are moved to the first rebooted compute node which doesn't have any vms, so the vxlan tunnels are re-created when the vms are evacuated to there. The vxlan tunnel should include one which is linked to the controller, and other vxlan tunnels linked to other compute nodes. However, sometimes only the vxlan tunnel(s) to other compute node(s) is/are created but the vxlan tunnel linked to the controller (where the open vswitch agent is active) is missing.

Environment

  • Red Hat OpenStack 6.0
  • openstack-neutron < 2014.2.3-32
  • neutron configured with l2population

Subscriber exclusive content

A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.

Current Customers and Partners

Log in for full access

Log In

New to Red Hat?

Learn more about Red Hat subscriptions

Using a Red Hat product through a public cloud?

How to access this content