Summary:	ASTERISK-02826: [request + patch?] SIP REGISTER timeout setting wanted
Reporter:	mack_jpn (mack_jpn)	Labels:
Date Opened:	2004-11-16 00:55:12.000-0600	Date Closed:	2008-01-15 15:16:24.000-0600
Priority:	Minor	Regression?	No
Status:	Closed/Complete	Components:	Core/General
Versions:		Frequency of Occurrence
Related Issues:
Environment:		Attachments:	( 0) chan_sip.regto.patch.txt ( 1) chan_sip.regto.patch2.txt ( 2) sip.conf.sample.patch.txt
Description:	Now I know we can change 20*1000 parameter for register interval time. r->timeout = ast_sched_add(sched, 20*1000, sip_reg_timeout, r); That is meaning send register at next 20sec later. if the parameter change 300*1000.send register at 5mins later.that is good for some carrier. Does anyone can implement 'register =>' line in sip.conf ?
Comments:	By: Brian West (bkw918) 2004-11-16 01:40:53.000-0600 ;maxexpirey=3600                ; Max length of incoming registration we allow ;defaultexpirey=120             ; Default length of incoming/outgoing registration Read the sip.conf.sample By: mack_jpn (mack_jpn) 2004-11-16 02:04:15.000-0600 nonono... I know,those parameters are for register with OK. I with to change the interval time for register with BAD condition. By: Brian West (bkw918) 2004-11-16 11:03:53.000-0600 backoff the register if it happens too fast?  IE if someone tries to brute force your boxen? bkw By: Olle Johansson (oej) 2004-11-16 11:26:21.000-0600 No, setting the timeout if we do not get a reply from the server... By: Olle Johansson (oej) 2004-11-16 11:26:40.000-0600 What does the RFC say about this? By: mack_jpn (mack_jpn) 2004-11-16 11:46:39.000-0600 Some japanese carrier says this situation is like a brute force. That isn't agree to them :-( By: khb (khb) 2004-11-16 15:41:52.000-0600 Yes, I also have a "registrationtimeout" parameter in my sip. See patch attached, but hasn't been compiled with current CVS. By: Brian West (bkw918) 2004-11-16 15:45:06.000-0600 but what does the RFC say? By: khb (khb) 2004-11-16 16:00:32.000-0600 It should be a matter of local policy, not an RFC concern. We are not in a dialog here and a 4xx is a final response. RFC says: 21.4 Request Failure 4xx   4xx responses are definite failure responses from a particular   server.  The client SHOULD NOT retry the same request without   modification (for example, adding appropriate authorization).   However, the same request to a different server might be successful. By: khb (khb) 2004-11-17 10:03:44.000-0600 config parameter should probably better be called "registertimeout" rather than "registrationtimeout" to indicate it's modifying "register=foo..." behavior. By: Brian West (bkw918) 2004-11-30 00:14:33.000-0600 Any updates.. pleas respond in the next 24 hours. bkw By: mack_jpn (mack_jpn) 2004-11-30 00:27:29.000-0600 hi khb ok. I agree. Thanks all. :-) By: khb (khb) 2004-11-30 00:36:00.000-0600 Updated patch.  changed option name to 'registertimeout' Disclaimed. Added sample configuration update. edited on: 11-30-04 00:46 By: Mark Spencer (markster) 2004-12-11 22:57:10.000-0600 Added to CVS, thanks! By: Russell Bryant (russell) 2004-12-15 14:16:44.000-0600 not in 1.0 By: Digium Subversion (svnbot) 2008-01-15 15:16:24.000-0600 Repository: asterisk Revision: 4434 U   trunk/channels/chan_sip.c U   trunk/configs/sip.conf.sample ------------------------------------------------------------------------ r4434 | markster | 2008-01-15 15:16:24 -0600 (Tue, 15 Jan 2008) | 2 lines Make registertimeout be configurable (bug ASTERISK-2826) ------------------------------------------------------------------------ http://svn.digium.com/view/asterisk?view=rev&revision=4434