sr-users

sr-users@lists.kamailio.org

27 participants
37381 discussions

Kamailio behind TLS-TCP load balancer
by David Villasmil 17 Aug '23

17 Aug '23

Hello all, I’m having lots of problems when trying to configure Kamailio behind an AWS tls load balancer to offload tls and receive on tcp on Kamailio. Everything else inside is UDP. I found I need to manually add record-route presets every time and invite comes in. And when trying to forward an ACK to the client via tls/tcp load balancer Kamailio complaint the socket is not TLS so it fails. Is there a simpler way of doing this via some parameters I don’t know? Thanks for helping me with this! David -- Regards, David Villasmil email: david.villasmil.work(a)gmail.com phone: +34669448337

5 9

kamailio-postgres on RHEL 9
by Ali Taher 17 Aug '23

17 Aug '23

Hello, I'm trying to install latest Kamailio version on RHEL 9.2. All is good, but I couldn't find kamailio-postgres package related to postgres module, knowing that I have installed Postgresql 15. Is there a rpm file that I can download to install kamailio-postgres ? Regards,

3 2

siremis Config.xml
by Maryam Baghdadi 15 Aug '23

15 Aug '23

Hello every one, How to set Remote Mysql database login ssl-mode in Config.xml file for siremis ???

1 0

Kamailio dns-failover / dns-loadbalancing with slow responding client.
by Mattis Lind 14 Aug '23

14 Aug '23

Hello! I am looking into a problem where we have Kamailio forwarding calls to two or more "recording-clients". I will try my best to describe the problem and would appreciate it if someone has an idea what to do. Please feel free to ask if you think I have forgotten to describe something that might be important or something is unclear in what I have written. We use use_dns_failover=yes and dns_srv_lb=yes so calls get load balancing to the "recording-clients". There is also the t_set_fr(60000,1000) parameter set so that if there is no response within 1 second it would try the next recording-client. The SRV record points to two or more recording clients. It now happens that the recording-clients sometimes have some kind of temporary problem so it will respond with a 503 after 5 seconds. What happens is that after the 1 second timeout trying to get the INVITE through to the first recording-client Kamailio will internally generate a 408. This will cause it to failover to another recording-client which happily takes care of the INVITE and responds properly with a 200 OK. Everything would have been just fine except for the fact that the first recording-client is just slow and finally responds with a 503. This 503 is not relayed backwards since a 200 has already been forwarded back to the caller. But when receiving the 503 Kamailio will initiate a new INVITE which is trying to set up a new call to a recording client. It looks like it is doing a new failover regardless if it already has done a failover for this failed transaction. I don't want Kamailio to send that last INVITE when receiving the 503. How can I configure Kamailio to disregard the last 503 (except for responding with an ACK) and not initiate a new INVITE? I have tried a lot of different changes to the configuration but failed to achieve this, unfortunately. Do I need to use the dispatcher module to achieve this? /Mattis

4 6

Modify To URI in Multiple Choices
by Ali Taher 11 Aug '23

11 Aug '23

Hello, I'm using Kamailio as SIP redirect where SBC sends INVITE packet to kamailio and "302","Multiple Choices" is sent back to SBC. What I want is to modify the URI part in the To header of the Multiple Choices packet sent from kamailio. I tried to use uac_replace_to("","sip:$var(cdpn)@$td") where $var(cdpn) is already defined in the routing logic, but it's not reflected and the original TO URI is sent instead. Any help to handle this issue is much appreciated. Regards,

2 2

how to get value of next hop
by Sergey Safarov 10 Aug '23

10 Aug '23

I receiving INVITE with two Route header (no to tag). First Route header is my Kamailio host. Second Route header is next hop after Kamailio. I need to get value of next hop. I tried to use "$route_uri" after "loose_route" but his deos works. "$route_uri" always use top Route header. But in the INVITE first Route header is Kamailio host. Is any change to get top Route header after "loose_route"?

3 2

Kamailio-asterisk BYE
by Dr. Barabás Péter 09 Aug '23

09 Aug '23

Hello, we work on a mobile softphone which connects to asterisk through kamailio. Almost everything is fine except one thing. Asterisk allows only one registration to an AOR. In mobile environment network can change, register/unregister can be called frequently in foreground/background cases. Kamailio forwards every request to asterisk. Until I have only one connection between kamailio – asterisk, everything is fine. The next situation is problematic: 1. Client A and client B register to asterisk through kamailio. –> there will be one connection between kamailio and asterisk 2. Client A calls client B. • one connection exist still 3. If client A terminates the call -> one connection remains. It is okay. 4. If client B terminates the call -> a new connection appears (kamailio IP with new port) After case 4 If I try to register again, I got 403 from asterisk, since the setting of asterisk does not allowed to register from different IP:port for the same AOR. (this behaviour cannot be changed now). The question: is there any setting/config exist in kamailio not to create new connection for a BYE in above case. Thanks in advance for your reply. Peter

1 0

Htable question
by Stefan-Cristian Mititelu 09 Aug '23

09 Aug '23

Hello, Can one kama process iterate a htable using sht_iterator_start(), sht_iterator_next(), sht_iterator_rm(), sht_iterator_end() while some other kama process call sht_rm(htable, item) ? I've looked into code and to me it looks like slots are locked ok. The entire linked list(specific to a slot) where the item might be, is locked, so this means there should be no problems, right? Thank you, Stefan

1 0

problem relay re-INVITE
by beer Ll 09 Aug '23

09 Aug '23

Hi I have a kamailio 5.7.1 running as TLS gateway for PBXs in private LAN network: UAC-WAN(sip-tls) <--> kamailio-WAN(sip-tls) - kamailio-LAN(sip) <--> PBX-LAN(sip) 80.0.0.1 <--> 80.0.0.2 - 10.0.0.1 <--> 10.0.0.2 All communications tested works well except for re-INVITE like when UAC put the call on hold . The re-INVITE packet is managed like a new call and the tm module doesn't match transaction : matching_3261(): RFC3261 transaction matching failed - via branch [z9hG4bK-524287-1---5e900485d5ad0faf] t_lookup_request(): no transaction found build_cell(): created new cell t_relay_to(): new INVITE I try to send the interesting logs and config call logs and pcap : https://filebin.net/vdezuhyn7wdy3lqc kamailio config : ... request_route { if (is_method("CANCEL")) { if (t_check_trans()) { route(RELAY); } exit; } if (!is_method("ACK")) { if(t_precheck_trans()) { t_check_trans(); exit; } t_check_trans(); } route(WITHINDLG); if ($si=="PBXIP0" || $si=="PBXIP1" || $si=="PBXIP2" || $si=="PBXIP3" ) { if (!loose_route()) { switch($rc) { case -2: sl_send_reply("403", "Forbidden"); exit; } } remove_hf("Route"); force_rport(); record_route(); $fs = "tls:WANIP:5061"; } if (is_method("INVITE|SUBSCRIBE")){ record_route(); } t_on_failure("MANAGE_FAILURE"); $fs = "udp:LANIP:5060"; } route(RELAY); exit; } route[WITHINDLG] { if (!has_totag()) return; if ($si=="PBXIP0" || $si=="PBXIP1" || $si=="PBXIP2" || $si=="PBXIP3") { route(RTPMANAGE); route(RELAY); exit; } if (loose_route()) { if ( is_method("NOTIFY") ) { record_route(); } route(RTPMANAGE); route(RELAY); exit; } if ( is_method("ACK|BYE") ) { route(RTPMANAGE); route(RELAY); } sl_send_reply("404","Not here"); exit; } How can I solve this ? Best Regards Leo

1 0

No
by Sergey Safarov 09 Aug '23

09 Aug '23

I catched issue with In the Kamailio logs i see ssl bug #1491 workaround: not enough memory for safe operation: shm=10219080 threshold2=11796480 Current stat about shared memory [root@sbc-a1 ~]# kamctl stats shmem { "jsonrpc": "2.0", "result": [ "shmem:fragments = 528", "shmem:free_size = 11343104", "shmem:max_used_size = 124449952", "shmem:real_used_size = 122874624", "shmem:total_size = 134217728", "shmem:used_size = 58728144" ], "id": 703182 } But when kamailio started [root@sbc-a1 ~]# kamctl stats shmem { "jsonrpc": "2.0", "result": [ "shmem:fragments = 122", "shmem:free_size = 112534264", "shmem:max_used_size = 22372528", "shmem:real_used_size = 21683464", "shmem:total_size = 134217728", "shmem:used_size = 12833888" ], "id": 703670 } What is stange a lot of fragments. How it can be troubleshooted? Can memory manager show info about each fragment like when fragment are created an and which module requested fragment?

2 3

Jump to page:

2025

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