Adam Carmichael wrote:

> I was considering relearning C/C++ all over again so I could help
> with this just because we need this feature quite badly. I'd love to
> finally submit something back to an opensource product, but I don't
> think my coding skills are quite up to scratch for this kind of
> challenge yet. I'd be more than happy to provide feedback and help
> with testing however, and I'll help patch what I can.

Thanks for the offer of assistance but I can manage the coding part
alone. Since the architecture of FR is really clean it's not a big
problem. And of course it much better if you do more tests on your
side when it's done.

> Users of this might need additional features, such as the ability to
> log other kinds of errors (such as RADIUS clients not in
> clients.conf (or the deprecated clients file) trying to
> authenticate, or if for example a particular NAS / LNS is running an
> old secret) so perhaps an auth_default_log() might also be required
> to log anything that doesn't match one of the other authentication
> types it could be logged into a kind of table that has a few BLOB or
> TEXT fields and places the entire log entry into that field. Another
> scenario is if you use ENCRYPT()'ed passwords within MySQL, and a
> user tries to log in using CHAP.

While writing the patch I gave up about the "auth_badpass_table" and
the "auth_goodpass_table". Just one "authlog_table" is sufficient, and
you chose what you put inside with the "authlog_query".

> What do you think? (with regards to the above outlined mysql logging
> scenarios). If you'd like to further development, then I'll
> subscribe to freeradius-developers and help out where I can.

I think indeed we should follow the discussion in the freeradius-devel
mailing list.

-- 
Nicolas Baradakis

- 
List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html

Reply via email to