Folks
I have several remotely-located servers, donated by folks not all of whom are computer geeks, let-alone Linux aware. In earlier versions of Centos, I directed them to perform a minimal NetInstall (not too difficult to direct over the phone), and then issue two commands after the first boot. One was a 'curl' to download a script of mine, and the second was a dot-command to run it. A few reboots followed in close succession, and I directed them to logon as root, and issue a simple dot-command. The personal intervention ended usually within an hour, since I had ssh access to the system and root, and could complete the installation and tailoring completely remotely. I would only need their intervention in the case of problems. In many cases, I've not needed to invoke that help for a year or more.
This scheme worked well until I decided to add support of graphical desktops on the systems in Centos 7, as I had in Centos 6. Once those packages were installed, further non-intuitive on-site intervention was required at next reboot to approve licenses. Remote ssh logins didn't work until that was done.
These licenses needed to be approved on the local console. This is quite annoying and places more demands upon my remote friend. In some cases, the console and keyboard had been removed and the box had been placed in the closet.
I am seeking a way to reduce or eliminate this annoyance. The best would be to find some way to 'pre-approve' the license agreement in my customizing scripts. Failing that, it would be nice to be able to approve these scripts from a remote ssh logon. I am reluctant to deploy Centos 7 without GUI support (I use x2go sometimes), and am thus reluctant to deploy centos 7 at this point.
Advice and comments welcome
David San Francisco