- 404 error messages on both of the appliances.
- Services unregistered and one service is unavailable state.
- Machine provisioning stuck on "in progress" and time out after a long time.
vRA Appliances in a clustered configuration have to come up in the opposite order they went down in due to RabbitMQ component.
This was resolved by using the correct shutdown procedure.
The correct shutdown order in clustered vRA appliances:
- Shutdown Node-2
- Shutdown Node-1
- Start up Node-1
- Start up Node-2
This will correctly start and initialize the RabbitMQ component to properly register all components.