sr-users November 2023

sr-users@lists.kamailio.org

47 participants
53 discussions

Incompatibility of the DMQ htable protocol with older Kamailio versions

by Denys Pozniak

Hello! I noticed that there are differences in the DMQ htable replication protocol for the old version of Kamailio 5.2 and the new 5.7. For example, the new version uses action 2 when updating values and action 4 to remove. The old Kamailio version uses the values 1 and 3 respectively. Well, as a result, the htable data is not shared between such nodes, plus there are errors in the syslog of the old version: ERROR: htable [ht_dmq.c:328]: ht_dmq_handle_msg(): failed to replay action It would probably be worth introducing some kind of DMQ htable protocol versioning to maintain compatibility? And is there any option to make them work together (I wouldn’t really want to change the action number in sip/sdp/json on the fly)? Kamialio 5.2 {"action":1,"htname":"shared","cname":"user","type":2,"strval":"uri","mode":1} {"action":3,"htname":"shared","cname":"user","mode":0} Kamailio 5.7 {"action":2,"htname":"shared","cname":"user","type":2,"strval":"uri","mode":1} {"action":4,"htname":"shared","cname":"user","mode":0} -- BR, Denys Pozniak

5 months, 1 week

tps_storage_fill_contact(): no r-uri user

by Denys Pozniak

Hello! My topology is like this: PBX -> Kamailio Topos -> User When calling the user, PBX does not add the user part in the Contact URI, something like: Contact: <sip:10.10.10.10:5080;transport=udp> When the customer sends a reinvite, the topos turns off for this transaction and the sip partially breaks down with error: topos [tps_storage.c:319]: tps_storage_fill_contact(): no r-uri user topos [tps_storage.c:615]: tps_storage_record(): failed to store As far as I understand, the RFC allows the absence of the user part in the URI and this occurs quite often. I can provide the pcap file and debug log in a private message if this helps identify the essence of the problem. # kamailio -v version: kamailio 5.7.1 (x86_64/linux) 1cf389-dirty -- BR, Denys Pozniak

5 months, 1 week

Kamailio ver 5.7.3 - source installation warnings

by Muhammad Danish Moosa

Hi, I am planning to upgrade Kamailio to version 5.7.3 to manage some memory leaks issues. I am installing from source and came across these warnings, just wanted to be sure if they are safe to be ignored. cr_func.c: In function ‘set_next_domain_on_rule’: cr_func.c:160:2: warning: missing braces around initializer [-Wmissing-braces] pv_value_t val = {0}; ^ cr_func.c:160:2: warning: (near initialization for ‘val.rs’) [-Wmissing-braces] cr_func.c: In function ‘actually_rewrite’: cr_func.c:348:2: warning: missing braces around initializer [-Wmissing-braces] pv_value_t val = {0}; ^ cr_func.c:348:2: warning: (near initialization for ‘val.rs’) [-Wmissing-braces] cr_func.c: In function ‘ki_cr_load_user_carrier_helper’: cr_func.c:782:2: warning: missing braces around initializer [-Wmissing-braces] pv_value_t val = {0}; ^ -- LD (gcc) [M p_usrloc.so] p_usrloc.so CC (gcc) [M evrexec.so] evrexec_mod.o evrexec_mod.c:64:1: warning: missing braces around initializer [-Wmissing-braces] static evrexec_info_t _evrexec_info = {0}; ^ evrexec_mod.c:64:1: warning: (near initialization for ‘_evrexec_info.data’) [-Wmissing-braces] LD (gcc) [M evrexec.so] evrexec.so CC (gcc) [M avp.so] avp.o LD (gcc) [M avp.so] avp.so CC (gcc) [M acc_diameter.so] diam_message.o CC (gcc) [M acc_diameter.so] acc_diameter_mod.o CC (gcc) [M acc_diameter.so] diam_avp.o CC (gcc) [M acc_diameter.so] diam_tcp.o LD (gcc) [M acc_diameter.so] acc_diameter.so CC (gcc) [M dmq.so] bind_dmq.o CC (gcc) [M dmq.so] message.o CC (gcc) [M dmq.so] dmq_funcs.o dmq_funcs.c: In function ‘ki_dmq_send_message’: dmq_funcs.c:304:3: warning: missing braces around initializer [-Wmissing-braces] dmq_peer_t new_peer = {0}; ^ dmq_funcs.c:304:3: warning: (near initialization for ‘new_peer.peer_id’) [-Wmissing-braces] dmq_funcs.c: In function ‘ki_dmq_bcast_message’: dmq_funcs.c:374:3: warning: missing braces around initializer [-Wmissing-braces] dmq_peer_t new_peer = {0}; ^ dmq_funcs.c:374:3: warning: (near initialization for ‘new_peer.peer_id’) [-Wmissing-braces] CC (gcc) [M dmq.so] notification_peer.o CC (gcc) [M dmq.so] dmq.o dmq.c:61:1: warning: missing braces around initializer [-Wmissing-braces] sip_uri_t dmq_server_uri = {0}; ^ dmq.c:61:1: warning: (near initialization for ‘dmq_server_uri.user’) [-Wmissing-braces] dmq.c:67:1: warning: missing braces around initializer [-Wmissing-braces] static sip_uri_t dmq_notification_uri = {0}; ^ dmq.c:67:1: warning: (near initialization for ‘dmq_notification_uri.user’) [-Wmissing-braces] Muhammad Danish Moosa " The core of mans' spirit comes from new experiences. "___ Christopher McCandless

5 months, 1 week

Info: Issue and PR automatic management

by Daniel-Constantin Mierla

Hello, during the Kamailio Development Meeting that took place in Dusseldorf earlier this month, one topic was related to administrative tasks related to project development and management, how to simplify/automatize such tasks. To reduce the work load on volunteering contributors, GitHub Actions were already used for various tasks related to project development and management (e.g., automatic builds on commits and pull requests to detect compile errors or code formatting mistakes). In Dusseldorf another task was configured to be managed with GitHub Actions, respectively the check of open issues and pull requests to evaluate the interest of submitters, developers and community users to pursue them. If there is no activity on an issue (potential bug or feature request) or a pull request, after 6 weeks it is marked with the label `stale`. After two more weeks of no activity, the issue or the pull request is marked with the label `expired` and closed. Note that any comment postpones the expire timeline, being considered that there is interest in pursuing the issue or the pull request. Requests for features were already treated in this way: if nobody commits to implement it, it can be closed after one month, but it needed manual work and many were still kept open. Potential bug reports that become very old are hard to tackle if the source code changes or new major releases are out, they might even not be valid anymore. Anyhow, this automatic operations can be reverted if there is still interest in pursuing the specific issues or pull requests. A registered developer can remove labels and reopen a closed issue or pull request. The non-registered-developer contributors have to make a comment that includes the token `/notstale` to remove the label `stale` or includes the token `/notexpired` to reopen a closed item. This new kind of automatic task management might add a little inconvenience because one has to restate the interest from time to time for those items that could not be addressed. However, considering that Kamailio is an open source collaborative project, in order to be fair for those that volunteer to spend time and resources for development of Kamailio, also the users/submitters have to stay engaged, not just report and forget about. The process to automatize tasks related to Kamailio development and administration is work in progress. Everything can be adjusted based on feedback (e.g., time lines), feel free to suggest improvements or new solutions to make things easier for everyone within the project ecosystem. Cheers, Daniel -- Daniel-Constantin Mierla (@ asipto.com) twitter.com/miconda -- linkedin.com/in/miconda Kamailio Consultancy and Development Services Kamailio Advanced Training -- asipto.com

5 months, 1 week

How to remove record-route and via from a forwarded call?

by Benoit Panizzon

Hi List Two Kamailio Nodes situation. Node A: Routing Instance. Node B: Registrar Instance. An invite is sent from Node A to B. Customer registered on B is 'busy' as example. B initiates Call Forwarding by adding a Diversion Header and sending the Invite back to A with a new R-URI towards the CFB destination. The Invite sent from B to A still has Node B in it's Record-Route. So all subsequent SIP messages pass via B, which is not needed and cause issue with rtpengine running on B. I attempted to remove_record_route(); when sending the call back to A, but this causes: parse_headers(): bad header field [;lr;ftag=3910078620-883101371> It looks like only the beginning of the RR header is removed leaving back a competely broken line. What is the propper was to remove a node from RR and Via routing? Mit freundlichen Grüssen -Benoît Panizzon- -- I m p r o W a r e A G - Leiter Commerce Kunden ______________________________________________________ Zurlindenstrasse 29 Tel +41 61 826 93 00 CH-4133 Pratteln Fax +41 61 826 93 01 Schweiz Web http://www.imp.ch ______________________________________________________

5 months, 1 week

WEBRTC-to-SIP & Vice Versa

by Vishal Pai

Hello All Good Day Here I am trying to achieve something like this WSS to UDP Ext registartion which works well using https://github.com/havfo/WEBRTC-to-SIP/tree/master/etc Note here i am forwarding the registartion to asterisk pbx server and define WSS of kamailio Now for calls - Webrtc - SIP calls works - SIP - Webrtc calls not working Here I am getting the request of the call on kamailio but kamailio send 404 not found to Asterisk PBX Webrtc - Webrtc calls not working Can anyone guide me on what I am doing wrong? Thanks Vishal P

5 months, 2 weeks

refer-to to new INVITE

by Social Boh

Hello, I'd like to know if there is a way, from Kamailio side, take the refer-to header Kamailio receive in a REFER request and create a new INVITE whit this data; then bridge the call between the originating user and the new destination user. Thank you Regards -- --- I'm SoCIaL, MayBe

5 months, 2 weeks

Re: RADIUS - Can't use extra attributes

by Duarte Rocha

Hello Daniel, Thanks for the reply. The extra dictionary is already included otherwise an error is thrown and Kamailio doesn't start. However, i realized my mistake now and you were right. I was checking for the extra attributes in a tcpdump and on radius server and i didn't had the dictionary configured on both of them. The extra attributes are in the packet but weren't being listed. One more question. Kamailio doesn't have a way to do a totally custom RADIUS access-request packet right? (Sorry if this opens another thread. I usually just hit reply on the mailing list and let Outlook create the e-mail, but it usually creates another thread and i'm not sure why). Thanks again. Cheers, Duarte

5 months, 2 weeks

Re: RADIUS - Can't use extra attributes

by Duarte Rocha

5 months, 2 weeks

Max-Forwards header missing in OPTIONS keep-alive messages

by Christian Schnell

Hello, in RFC 3261 (sec. 8.1.1.6) it states about the Max-Forwards header field: > A UAC MUST insert a Max-Forwards header field into each request it originates with a value that SHOULD be 70. > > Source: https://datatracker.ietf.org/doc/html/rfc3261#section-8.1.1.6 When Kamailio sends its OPTIONS keep-alive requests (from the nathelper module I believe), Max-Forwards is missing: > OPTIONS sip:alice@example.com:61266 SIP/2.0 > Via: SIP/2.0/UDP aa.bb.cc.dd:5060;branch=z9hG4bK5124450 > From: sip:pinger@sip.example.com;tag=uloc-6550aa14-7456-2461-6390f173-5fdb4667 > To: sip:alice@example.com:61266 > Call-ID: aba25282-2ca72be1-9d1fd53(a)aa.bb.cc.dd > CSeq: 1 OPTIONS > Content-Length: 0 The user agent library I'm using (libre from baresip) rejects these requests with a "400 Bad Request" response, I looked at the sources and it denies these requests because of the missing Max-Forwards header. Of course this suffices for the keep-alive logic because all that's needed is any traffic in both directions to keep both (potential) NATs open. However you might agree that it does look a bit confusing at first, and the RFC is clear about that. So it's not a bug, but is this behavior intended? Thanks, also thanks for all the great work that has been put into Kamailio. Christian

5 months, 2 weeks

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

sr-users November 2023