Ok, I don't see why the Ubuntu tests fail. On the first glance, this looks unrelated.

Yeah it's an upstream bug ansible/ansible#23530. Also geerlingguy/drupal-vm#1302.

Unfortunately I don't think the more ideal fix of {{ ansible_user|default(ansible_user_id) }} will work on Ubuntu 12.04 but maybe worth a try and see what travis says. If the variable defaulted to ansible_user_id, the docs might need to highlight even more that the user should be changed to a non-root user account, as that's the default otherwise.

Ah, thanks for the explanation and pointing this out. Then it is in fact unrelated and further action concerning this is out of scope of this PR.

Is there anything I can do to help getting this merged?

I believe the fix is geerlingguy/ansible-role-backup@72b9945

Ok, I incorporated this change. Unfortunately, it seems CentOS 6 still can't find the variable.

I think this will work:

certbot_auto_renew_user: "{{ lookup('env', 'USER') }}"

Finally :D thanks @tterranigma

Hopefully it's ready to merge now @geerlingguy

I rebased and cleaned up this PR. While this has nothing directly to do with using the staging server, it generally allows deploying the /etc/letsencrypt/cli.ini configuration file with whatever options one wants.

@geerlingguy do you want this to go in? If yes feel free to merge, but also feel free to close otherwise. Meanwhile I'm using an own role tailored to the needs of my infrastructure, so this is no longer important to me but I wanted to keep the contribution usable and I'm interested in resolving this PR.

could this be merged @geerlingguy ?

closed since it has conflicts and I'm no longer maintaining this PR