Fantastic!
> 1) Am I up to date wrt to patches with the fixes to the syslog and HUP bugs?
Not to be flip, but "I don't know, are you up to date?" is what comes to
mind. I apologize for these errors I introduced, and while I've been on
vacation I've tried to help people through this mess, including posting
some diffs. When I get back, I plan to place the diffs on ftp.merit.edu
and work on a clean release. If you have changed lines 587 and 3748 of
the radiusd.c file (compared to the 2.4 distribution) then you have them.
> 2) Is there more complete documentation on the AATV API? The documentation
> that I have seen so far seems rather sketchy.
Gee, I thought it was pretty good. :-) But then I'm not a technical
documentation type. This is the best there is, you can ask me questions
and from those questions and from people's experience I'll update the
aatv.cookbook.txt and aatv.txt files. A new release like this is bound
to have rough edges IMHO...
> 3) It appears that the AATV API will handle accounting as well. Is there
> any documentation on how that differs from using the API for authorization?
The AATV API is meant to be quite general and is today handling accounting
in production systems within MichNet. There isn't any such docoumentation
because it doesn't differ. Maybe I missed something in your question...
> 4) It appears that there is a built-in method to prevent more than one
> simultanous login. Is this correct? If so, how is it used?
Yes. This is what I plan to release next week -- I don't recall what we
had in the release and what is yet to release -- being on vacation tends
to dull one's keen, technical edge... ;^)
> 5) Is there any way to prevent the radius server from timing out and exiting
> other than editing the code? I'm aware of the -t parameter, but that
> doesn't seem to do what I want it to.
The Merit version has been designed that way for as long as I can remember,
well, nearly two years, anyway... It is designed to run under the daemon
inetd or xinetd and to go away because of inactivity after a compile-time
constant (DEFAULT_INETD_TIMEOUT) number of minutes. I guess you could use
the "-t" command line option to override that with a very large number of
minutes (effectively approaching infinity...)
Regards,
web...
-- William Bulley, N8NXN Senior Systems Research Programmer Merit Network Inc. Domain: web@merit.edu 4251 Plymouth Road MaBell: (313) 764-9993 Ann Arbor, Michigan 48105-2785 Fax: (313) 747-3185