Re: rlm_sql.c and sql.c fixes for Simultaneous-Use
Thank you, it works like a charm now! Sorry for the wasted effort yesterday with the patches to sql.c, etc. Chris Kalin - Original Message - From: Kostas Kalevras [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Monday, August 26, 2002 8:00 PM Subject: Re: rlm_sql.c and sql.c fixes for Simultaneous-Use On Mon, 26 Aug 2002, Chris A. Kalin wrote: OK, I'll buy that. :) The problem is when it calls rad_check_ts, _that_ function seems to be checking a radutmp file - whether or not I'm using radutmp or not for session checking - and several variables come back undefined, and hence checkrad still doesn't get fired up. So between the two, this seemed a more expedient solution to my particular problem (I'm only using sql, not radutmp). I'll dig over in rad_check_ts and go from there. Thanks for the help, at least eliminating sql_check_ts from this makes the troubleshooting a little more obvious. Chris Kalin rad_check_ts will just call checkrad. Please do a cvs update and check if it works better now. rad_check_ts should now be called with the proper arguments in rlm_sql_checksimul. -- Kostas Kalevras Network Operations Center [EMAIL PROTECTED] National Technical University of Athens, Greece Work Phone: +30 10 7721861 'Go back to the shadow' Gandalf - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
rlm_sql.c and sql.c fixes for Simultaneous-Use
I've discovered a few little quirks in rlm_sql.c and sql.c that prevent Simultaneous-Use from starting radcheck in a standard FreeRADIUS install. Without these fixes, Simultaneous-Use will still work in a trust the SQL server context, but it's unable to verify the information using radcheck - which can bite people if RADIUS Stop packets have been dropped, etc. It's a 3 KB context patch or a 1 KB diff file. May I post it here or is there a more appropriate venue? Thanks, Chris Kalin Netwurx, Inc. - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
Re: rlm_sql.c and sql.c fixes for Simultaneous-Use
At 03:51 PM 8/26/2002 -0500, Chris A. Kalin wrote: I've discovered a few little quirks in rlm_sql.c and sql.c that prevent Simultaneous-Use from starting radcheck in a standard FreeRADIUS install. Without these fixes, Simultaneous-Use will still work in a trust the SQL server context, but it's unable to verify the information using radcheck - which can bite people if RADIUS Stop packets have been dropped, etc. It's a 3 KB context patch or a 1 KB diff file. May I post it here or is there a more appropriate venue? Post the patch on a webpage, and post the link here. We'll take a look at it and review it. If it fits the development criteria, we'll incorporate it. Thanks! -Chris -- \\\|||/// \ StarNet Inc. \ Chris Parker \ ~ ~ / \ WX *is* Wireless!\ Director, Engineering | |\ http://www.starnetwx.net \ (847) 963-0116 oOo---(_)---oOo--\-- \ Wholesale Internet Services - http://www.megapop.net - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
Re: rlm_sql.c and sql.c fixes for Simultaneous-Use
Chris Parker wrote: At 03:51 PM 8/26/2002 -0500, Chris A. Kalin wrote: I've discovered a few little quirks in rlm_sql.c and sql.c that prevent Simultaneous-Use from starting radcheck in a standard FreeRADIUS install. Without these fixes, Simultaneous-Use will still work in a trust the SQL server context, but it's unable to verify the information using radcheck - which can bite people if RADIUS Stop packets have been dropped, etc. It's a 3 KB context patch or a 1 KB diff file. May I post it here or is there a more appropriate venue? Post the patch on a webpage, and post the link here. We'll take a look at it and review it. If it fits the development criteria, we'll incorporate it. OK, the diff is at http://www.netwurx.net/~cak/patch-sql.diff and the context diff is at http://www.netwurx.net/~cak/patch-sql.context . Should be simple enough to understand without much comment. It's against last night's CVS snapshot. Chris Kalin - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
Re: rlm_sql.c and sql.c fixes for Simultaneous-Use
Chris A. Kalin [EMAIL PROTECTED] wrote: Should be simple enough to understand without much comment. It's against last night's CVS snapshot. OK... but you changed it so that it doesn't call rad_check_ts() any more, and instead calls sql_check_ts(). This is wrong. The issue is that SQL may not know that the user is logged in twice, but another module may know. So hard-coding the SQL module to only check the SQL functions, is wrong. The sql_check_ts() function should be deleted. See: http://lists.cistron.nl/pipermail/freeradius-devel/2002-August/003376.html Coincidentally enough, also from today. Alan DeKok. - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
Re: rlm_sql.c and sql.c fixes for Simultaneous-Use
OK, I'll buy that. :) The problem is when it calls rad_check_ts, _that_ function seems to be checking a radutmp file - whether or not I'm using radutmp or not for session checking - and several variables come back undefined, and hence checkrad still doesn't get fired up. So between the two, this seemed a more expedient solution to my particular problem (I'm only using sql, not radutmp). I'll dig over in rad_check_ts and go from there. Thanks for the help, at least eliminating sql_check_ts from this makes the troubleshooting a little more obvious. Chris Kalin - Original Message - From: Alan DeKok [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Monday, August 26, 2002 4:50 PM Subject: Re: rlm_sql.c and sql.c fixes for Simultaneous-Use Chris A. Kalin [EMAIL PROTECTED] wrote: Should be simple enough to understand without much comment. It's against last night's CVS snapshot. OK... but you changed it so that it doesn't call rad_check_ts() any more, and instead calls sql_check_ts(). This is wrong. The issue is that SQL may not know that the user is logged in twice, but another module may know. So hard-coding the SQL module to only check the SQL functions, is wrong. The sql_check_ts() function should be deleted. See: http://lists.cistron.nl/pipermail/freeradius-devel/2002-August/003376.html Coincidentally enough, also from today. 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
