Ansible doesn’t mirror existing systems you have out in the field. Before you go further, I’ll try to convince you that you don’t actually want what you are asking for.
What files do you consider important? What SELinux attributes? What configurations? Cloning an existing system is basically the “golden image” kind of workflow that the DevOps community seeks to get away from.
What data in the database is important?
For, really, when you want to make changes, where do you make them? Do you go back to this trusted system forever? You don’t want that.
What it does is configure remote systems based on your orders/configuration/policy to make them easily reproducible and cloneable.
This can include the base system, the software deployed on top of it, even the processes by which multiple machines in a cluster are deployed in unison.
It can also be run in “check mode” to tell you if a given machine is likely in compliance with your established policy.
I’m not sure what you mean about OpenVZ, CentOS, and custom OS info.
Definitely a configuration management tool is not going to make it completely easier to help you switch operating systems – for instance, think about the how the Apache structure is different between Ubuntu and CentOS, or how you don’t have SELinux on Ubuntu, but it will definitely help
making you more adaptable.