Due to a current issue with RHEL 7.5, the following procedure must be run once on any compute hosts where SWe instances which require multi-queue virtio interfaces will be deployed. These steps enable such instances to spawn. The procedure confirms that the failure to spawn is due to the underlying RHEL 7.5 issue and then provides steps to work around this issue.
/var/log/audit
audit.log
file and confirm that the following denial is reported by SELinux:avc: denied { attach_queue }
audit2allow -a -M attach_queue
semodule -i attach_queue.pp
After you have prepared your OpenStack environment (refer to Pre-Instantiation Requirements of OpenStack) and developed a Heat template for your SBC SWe deployment (refer to Developing a Heat Template), you can use standard OpenStack methods to instantiate an SBC SWe instance. You can use OpenStack Heat commands or you can use the OpenStack Horizon dashboard to create the SBC instance. Refer to OpenStack documentation or documentation from your OpenStack provider for information on deployment using Heat.
Regardless of method, during orchestration you must specify the name of the template file you want to use to create your instance. If the template file does not provide an acceptable default value for any of the parameters specified within the template, then you must provide those values when you instantiate. You can provide these values in an environment file that you identify along with the template file when you instantiate. Otherwise, the values can be provided on the command line when you create the instance or entered in the GUI screen presented when you use the Horizon UI to launch an instance. The exact parameters you are required to provide or the fields that appear on a GUI screen depend on the parameters defined in your template.
Make sure all your NICs are physically plugged in (link light on) before spawning VM instances. Otherwise, an incorrect port mapping may occur (logical to physical), and your SBC may not function properly.
If you are instantiating an SBC cluster, you can view the status of the nodes in the EMS GUI after they register with the EMS:
Once an N:1 Redundancy Group (RG) is created, you must not delete any instances from the RG using the OpenStack dashboard GUI. In this case, you must remove the RG and create a new one.
If a switchover is triggered on the SBC SWe platform due to any of the following reasons, a media outage of 20 seconds occurs for the stable calls: