Hi OpenStack Community,

I am currently facing an issue with the Load Balancer (Octavia) service in my OpenStack deployment and would appreciate your guidance or suggestions on how to troubleshoot and resolve it.

Environment Details:

• OpenStack Version: 2023.1 (Antelope)
  

• Operating System: Ubuntu 22.04
  

• Octavia Backend: Amphora
  

• Network Type: VXLAN
  

• Deployment Tool: Kolla Ansible 
  

Issue Summary:

After creating a Load Balancer, the Amphora instance is successfully launched. However, approximately 18 minutes later, the Load Balancer transitions to an ERROR state.

Error Observed:

octavia.common.exceptions.ComputeWaitTimeoutException: Waiting for compute id a222527f-8336-4fb1-8e9e-0e3ac6e8d580 to go active timeout.
octavia.amphorae.driver_exceptions.exceptions.AmpConnectionRetry: Could not connect to amphora, exception caught:
HTTPSConnectionPool(host='192.168.200.247', port=9443): Max retries exceeded with url: // (Caused by NewConnectionError: Failed to establish a new connection: [Errno 113] No route to host)

What I’ve Tried:

1. Created a VXLAN-based internal network:
   

   openstack subnet create --network octavia-mgmt-net --subnet-range 192.168.200.0/24 octavia-mgmt-subnet
   openstack router create octavia-router
   openstack router add subnet octavia-router octavia-mgmt-subnet
   openstack router set octavia-router --external-gateway <public-network-id>
   
   

2. Manually configured OVS interface:
   

   ovs-vsctl add-port br-int octavia-mgmt -- set interface octavia-mgmt type=internal
   ip addr add 192.168.200.1/24 dev octavia-mgmt
   ip link set octavia-mgmt up
   
   

3. Security groups are open (allowing all ingress and egress).
   

4. Verified OVS bridge setup – interface appears correctly on br-int.
   

Behavior Noticed:

• Amphora instance gets created but fails to communicate with the controller after a while,so health checks fails.
  

• The controller cannot reach the Amphora management IP (192.168.200.x).
  

• Logs show heartbeat/connection issues (No route to host, timeouts).
  

Request for Help:

I would appreciate any of the following:

• Suggestions to verify if the management network is properly routed from controller to Amphora.
  

• Best practices for setting up the Octavia management network using VXLAN.
  

• Troubleshooting steps for verifying health manager connectivity to Amphora.