FreeRadius 3.0.10, radclient status failure

[Alan DeKok]

On Nov 2, 2015, at 9:31 PM, Arran Cudbard-Bell <a.cudbardb at freeradius.org> wrote:
> I believe the previous behaviour was correct,

  As the author of RFC 5997, I would suggest that the default is to expect Access-Accept.

>  but this'll help the people who can't read log output that explains in detail what the problem was and how to fix it.

  The problem is you don't have to specify attribute filters when sending Access-Request packets.  So the treatment of Status-Server in radclient is magical and unexpected...

  I suggest that having magical behaviour is wrong.

  I suggest that having defaults which work in most cases are correct and useful.

> Though it'll now fail in an obscure and annoying way (that doesn't explain exactly what the problem was an how to fix it) when sending Status-Server requests to a non standard accounting port.

  Or, we could fix radclient to not care what the response is to Status-Server.  That would be the better approach.

  Alan DeKok.