On Thu, 14 Nov 2002, Nathan Miller wrote: > I seem to be having the same issue.. seems to happen randomly about once a > week on a production server running a simple perl backend for authentication. > > snippet of log... > Thu Nov 14 15:26:38 2002 : Error: WARNING: Unresponsive child (id 65559) > for request 13464 > Thu Nov 14 15:26:38 2002 : Error: WARNING: Unresponsive child (id 66584) > for request 13465 > Thu Nov 14 15:26:39 2002 : Error: CHILD: exit on signal (11) > > Now this is a production server recently upgraded from v0.3 to > v0.7.1. v.0.3 did not crash like this. As soon as you get those errors > all instances of radius crash and burn. Let's just say if a external auth > process gets stuck, why is the entire radius server bombing and not just > that thread aborting? Seems goofy to me.
This is a known bug which has been fixed in the cvs head. Use the latest cvs snapshots and you should be fine. > > At 10:13 AM 10/30/2002 -0500, you wrote: > >Igor Chen <[EMAIL PROTECTED]> wrote: > > > The main reason of core dumping was delay after sending request to > > > database. Trigger on UPDATE became too slow (UPDATE request was handled > > > ~40 - 60 sec.) > > > > This problem should be addressed in the documentation in flaming > >letters 10 feet high. If the back-end database takes more than 5 > >seconds to respond to a request, then the RADIUS server will not be > >able to authenticate people. > > > > Once the database is broken, the RADIUS server (which depends on the > >database) can't be any better. > > > > Alan DeKok. > > > >- > >List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html > > > - > List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html > -- Kostas Kalevras Network Operations Center [EMAIL PROTECTED] National Technical University of Athens, Greece Work Phone: +30 210 7721861 'Go back to the shadow' Gandalf - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
