Hi,
I am relatively new to Ansible, but have been very impressed with its functionality and excited about it’s possibilities. I am primarily using it for OpenStack orchestration on a Nebula private cloud for launching Linux with various distributions. This has been very successful so far although I have run into some differences in the Nebula OpenStack implementation with regards to Modules. I plan on writing some custom modules for Nebula specific OpenStack automation, but in the meantime I am building my experience with using Ansible.
Can folks suggest ways to dynamically manage host inventories when launching cloud instances? Looking at the documentation there are ways of “registering” the IP address returned by a Nova command for instance (have done this successfully), but that information seems transient and is only available in the same playbook which really erodes the modularity of Ansible. Ideally I would be able to update an inventory list that other roles could use after I have instantiated my instances.
In short I am imagining roles for launching an instance and roles for other other things like web servers or database instances. But at the moment I am at a loss on how to use the IP addresses returned after launching an instance in a host inventory that another role can use without manual intervention.
Perhaps I have missed the point or missed other discussions on this.
Thanks,
Albion