Alan Buxey <alan.buxey@gmail.com> writes:
if the config is present and done, yes - as Alan has said, look at control-socket
another useful method is to have a 2nd instance of FR that listens on different ports but with the same main config etc as the production server (or just use another box or docker container etc) - then have a specific AP or switch configured to talk to only that FR instance and put your client on that. its like a little medical treatment room to see WTF the client issue is :-)
I recommend anyone who hasn't yet tried it to take a deep look at the control-socket. Play with it for a while so you know what it can do. It's an incredibly powerful tool. I just love the ability to conditionally debug specific requests on a busy production server. Any production NAS would quickly produce a lot of debugging noise on your 2nd instance. The control socket let you debug just that single user with issues, or a single Calling-Station-Id or whatever conditional debug filter you want. Without touching the NAS config at all. It's just beautiful :-) But do read the the warnings in the example config. You obviously need to trust anyone with access to the control socket. Shouldn't be a big problem, since you really shouldn't allow anyone you don't trust to access your radius servers at all. Bjørn