Hey!

Yes, I think this should be pretty much correct. Though, ideally, you should not need the /etc/ansible/ directory - it should be populated as part of the bootstrap process. Also /opt/openstack-ansible should usually be just a result of git clone and git checkout, as no manual edits should be done inside of it.

So the only thing you actually need to have for such cases, is /etc/openstack_deploy directory, which contains all configuration for OpenStack-Ansible.

But yeah, this looks good to me, and I believe it should just work :)




чт, 25 сент. 2025 г. в 15:17, Danish Khan <danish52.jmi@gmail.com>:
Dear Team,

Our controller1 node which was also openstack deployment node somehow crashed and is not accessible now.

I am trying to make controller3 as new deployment node. 

Below is the step which I am planning to follow. I have backup of (openstack-ansible, openstack_deploy and ansible directory of controller1 ):

1. Place /opt/openstack-ansible directory from controller1 to controller3
2. copy /etc/openstack_deploy directory from controller1 to controller3
3. copy /etc/ansible/ directory from controller1 to controller3

4. run bootstrap on controller3 
5. test if openstack-ansible has picked same clusters and inventory.


Can someone please suggest if this is correct approach or if I am missing something.

Regard,
Danish