Assertion failed in request_list.c

[Alexander Voropay]

Hi!

 I still have a problem described there :

http://lists.freeradius.org/pipermail/freeradius-devel/2005-August/008670.html

(see a thread)

 The FreeRadius sources is from the CVS a week ago.

 I can't use FreeRADIUS on the heavy load at all...  It dies **every** 2..3 min :(
There is no such problem under lite load (upto ~1req/ per sec.).


Fri Sep 16 12:10:51 MSD 2005: Radius died, restarting..
/opt/radiusd_credit/sbin/radwatch: line 50:  8419 Aborted                 $RADIUSD -f $*
Fri Sep 16 12:11:47 MSD 2005: Radius died, restarting..
/opt/radiusd_credit/sbin/radwatch: line 50:  8444 Aborted                 $RADIUSD -f $*
Fri Sep 16 12:15:16 MSD 2005: Radius died, restarting..
/opt/radiusd_credit/sbin/radwatch: line 50:  8471 Aborted                 $RADIUSD -f $*


Fri Sep 16 12:10:51 2005 : Error: Assertion failed in request_list.c, line 577
Fri Sep 16 12:11:01 2005 : Info: Ready to process requests.
Fri Sep 16 12:11:01 2005 : Auth: Login OK: [user] (from client c5850a port 1144 cli 2222222)


 I've increased a number of severs in the radiusd.conf, no changes.

 Is there any way to avoid this problem ?

 My platform is Linux/PPC64 so I can't use valgrind tool to debug a problem.
Is it possible to add more information

 It seems there is some problems with main request queue list. May be,
some atomic locking around list access will save us ?

 Is it possible to change multithread model insted of the NPTL ? 

 Could you add some debug info (request queue status/statistics)
in the 'assert' ?


 I've noticed one static definition of the array :

=== src/main/request_list.c: 57 ========
struct request_list_t {
        REQUESTINFO     request_list[256];
===========

 May be, it exceeds a limit on the heavy load ?

--
-=AV=-