Ansible Control Machine failure and long running tasks

[Nicolas B]

Hello!

Recently, I have started looking at Ansible to be used as an orchestration engine. I need the engine to support long running tasks, and as Ansible supports asynchronous actions with polling, it could fit my use case perfectly.

However, I was wondering what would happen if the control machine encountered a failure? (like an unexpected shutdown, network outage, etc.) From what I read, Tower provides high availability features with one active node and some others passive nodes ready to take over if needed, but does it work in the middle of a playbook?

Thanks in advance for any help!

Nicolas