AWX how to deploy, scale and use execution node, hybrid node and hop node.

Hi,

I have been looking for a while for anything on the topic of how to actually provision additional execution instances, using instance groups, EE etc. and I’m still unclear on how to achieve this with k8s and awx-operator.

Long story short I’m evaluating use of AWX to manage some of my cloud envs. I understand that there is a concept of receptors to provide a mesh network which I’m hoping can help reaching nodes behind NAT etc. However I have zero idea how to actually go about it given current AWX setup.

Does anyone has any experience with this or similar topics and can point my to some reading/viewing materials or even some woking configs/k8s manifests etc. ?

I though this would be something that can be done from GUI but it does not seem to be the case, unless I’m missing something.

I can see on the topology view legend has control plane, execution nodes, hybrid node and hop nodes but I have no clue how to actually deploy and use those.

Any help would be greatly appreciated.

Thank you

Hi. Short answer is you can’t. AWX deployments are k8s based so you can use container groups for extra automation capacity.

Can I have the long answer :wink:

If one would to edit the receptor conf (via a volume mount) and set up a receptor network and do the awx manage commands. Would it not work? .

(Same way the dev docker compose launches a hop node)

For outside k8s you would need to define a recepor service.

I do agree that this is not standard awx but should it not work if someone would self manage the integration of receptor and awx?

Is there some technical obstacle I am missing?

Kind regard,