sr-users December 2025

sr-users@lists.kamailio.org

28 participants
25 discussions

Replacing an attribute in the SDP
by David Cunningham 19 Dec '25

19 Dec '25

Hello, We need to replace a=sendrecv with a=sendonly in the 200 OK reply to an INVITE. In theory it should be something straightforward, like in the onreply_route using: replace_body_all( "a=sendrecv", "a=sendonly" ); or: subst_body( '/^a=sendrecv/a=sendonly/' ) Our first problem is that these both are appending a=sendonly to the end of the SDP, not replacing values within it. Does anyone know why this might happen? Our second problem is that we really only want to do this for video media, and not touch the audio media. Are there any text tools that allow you to work with specific media within the SDP? Thanks in advance for any tips! -- David Cunningham, Voisonics Limited http://voisonics.com/ USA: +1 213 221 1092 New Zealand: +64 (0)28 2558 3782

3 6

Regresion from 5.8.4 to 5.8.6
by Ihor Olkhovskyi 18 Dec '25

18 Dec '25

Not sure where the bug is, just noticed when playing on local development Kamailio installation, that simple rewrite of $fu start to act funky. Initial From field in a packet is: From: sip:90001@localhost.localdomain ;tag=83017fca-9fe4-493e-9e66-406034482628 In the code I'm having a statement $fu = "sip:" + $avp(callerID) + "@localhost.localdomain"; Resulting From field on 5.8.4 looks like From: sip:90001@localhost.localdomain ;tag=83017fca-9fe4-493e-9e66-406034482628 Note the 2 spaces after From, but it's ok, I guess, no other changes which is fine. But exact the same config file on 5.8.6 giving this: From: <sip:90001@localhost.localdomain>sip:90001@localhost.localdomain ;tag=83017fca-9fe4-493e-9e66-406034482628 Which is not correct SIP URI. I'm not sure when the issue was introduced, but that's what I've found. If will have a bit more time, will try to investigate deeper, maybe just to warn about possible bugs with the versions greater than 5.8.4 (have not tested 5.8.5) -- Best regards, Ihor (Igor)

3 9

listen with advertise and a custom port
by Володимир Іванець 18 Dec '25

18 Dec '25

Hello everyone! I was testing a few infrastructure scenarios and found interesting behavior that may be not correct. So I decided to ask. I'm running a few day old Kamailio build from master branch. Listeners are: listen=udp:10.0.40.13:5060 advertise X.X.X.X:5080 name "sock1" listen=udp:10.42.3.1:5060 name "sock2" Replies to the incoming requests to sock1 are updated with X.X.X.X address but the custom port is not added. I tried changing the local port too but Record-Route and Contact headers did not have a custom port added. The only way I was able to force Kamailio to use a custom port was to make it a part of an address by including both in quotes: listen=udp:10.0.40.13:5060 advertise "X.X.X.X:5080" name "sock1" Is there a correct way to achieve this? Thanks a lot!

2 1

Re: How to add date/hour in Kamailio logs?
by Henning Westerholt 17 Dec '25

17 Dec '25

Hello, Please keep the list address in the mail discussion. in this case maybe just look to the logstash date filter plugin, it seems it also can add timestamps. You’ve mentioned before that you are using an ELK stack. Cheers, Henning From: Benedito Marques <benedito.marques1(a)gmail.com> Sent: Mittwoch, 17. Dezember 2025 13:58 To: Henning Westerholt <hw(a)gilawa.com> Subject: Re: [SR-Users] How to add date/hour in Kamailio logs? I'm using the kamailio application inside docker/k8s. Even in this way, its possible to use kamailio + rsyslog to print debug messages to stdout? On Tue, Dec 16, 2025 at 5:01 AM Henning Westerholt <hw(a)gilawa.com<mailto:hw@gilawa.com>> wrote: Hello, the log prefix will be not used for some error messages, as you correctly noticed. One easy workaround is to just use a logging daemon like rsyslog and add the timestamp there before you route it to later processing steps. Cheers, Henning -- Henning Westerholt – https://skalatan.de/blog/ Kamailio services – https://gilawa.com<https://gilawa.com/> From: Benedito Marques via sr-users <sr-users(a)lists.kamailio.org<mailto:sr-users@lists.kamailio.org>> Sent: Montag, 15. Dezember 2025 22:23 To: Kamailio (SER) - Users Mailing List <sr-users(a)lists.kamailio.org<mailto:sr-users@lists.kamailio.org>> Cc: Benedito Marques <benedito.marques1(a)gmail.com<mailto:benedito.marques1@gmail.com>> Subject: [SR-Users] How to add date/hour in Kamailio logs? It's possible to add date/hour or add more information in Kamailio logs? I'm using kamailio inside containers. Even have an option in container manager (--timestamps in kubectl/docker for example), it's not possible to get some kamailio date/hour logs when using external logs tools like ELK. I add this option to get date/hour in xlog: log_prefix="[$Tf] - " but some logs (like bellow) generated automatically by Kamailio have no this information: Listening on tcp: 172.20.0.4:5091<http://172.20.0.4:5091> name tcp_socket advertise 200.200.200.200:5091<http://200.200.200.200:5091> tls: 172.20.0.4:5063<http://172.20.0.4:5063> name tls_socket advertise 200.200.200.200:5063<http://200.200.200.200:5063> Aliases: tls: 2eab5fb54462:5063 tcp: 2eab5fb54462:5091 0(25) WARNING: tls [tls_init.c:807]: tls_h_mod_init_f(): openssl bug #1491 (crash/mem leaks on low memory) workaround enabled (on low memory tls operations will fail preemptively) with free memory thresholds 204800 and 102400 bytes 16(41) ERROR: tls [tls_server.c:1329]: tls_h_read_f(): protocol level error 16(41) ERROR: tls [tls_util.h:42]: tls_err_ret(): TLS accept:error:1408F09C:SSL routines:ssl3_get_record:http request 16(41) ERROR: tls [tls_server.c:1333]: tls_h_read_f(): source IP: 165.232.66.142 16(41) ERROR: tls [tls_server.c:1336]: tls_h_read_f(): destination IP: 172.20.0.4 16(41) ERROR: <core> [core/tcp_read.c:1479]: tcp_read_req(): ERROR: tcp_read_req: error reading - c: 0x7fc51458ef30 r: 0x7fc51458f058 (-1) -- Att, Benedito Marques -- Att, Benedito Marques

1 0

High volume ratelimit CPS
by Jon Bonilla (Manwe) 16 Dec '25

16 Dec '25

Hi all Some time ago I asked about a way of limiting the CAPS sent to differnet gateways and in high CPS environment. I ended up implementing a htable based solution suggested by one member of the list. The solution was more or less like this one: modparam("htable", "htable", "caps=>size=11;initval=0;autoexpire=5;") modparam("htable", "timer_interval", 10) modparam("htable", "timer_procs", 4) $var(cps) = $rd + ":" + $timef(%S); $var(cps) = $shtinc(caps=>$var(cps)); if ($var(cps) > $var(capslimit)) { sl_send_reply("503", "CAPS limit reached for this destination $rd"); exit; } The problem with this approach is that it's not fast enough detecting the limit. If I set 500 CAPS as limit it will sometimes reach 550, 600, or even 700 before it rejects the rest of the requests during that second. I have complains from the destination and they are asking me to be more precise in the ratelimit process. Would ratelimit/pipelimit improve this? I don't even know if they would accept handling 1 second intervals at this rate. Any other suggestion? cheers, Jon -- PekePBX, the multitenant PBX solution https://pekepbx.com

5 10

How to add date/hour in Kamailio logs?
by Benedito Marques 16 Dec '25

16 Dec '25

It's possible to add date/hour or add more information in Kamailio logs? I'm using kamailio inside containers. Even have an option in container manager (--timestamps in kubectl/docker for example), it's not possible to get some kamailio date/hour logs when using external logs tools like ELK. I add this option to get date/hour in xlog: log_prefix="[$Tf] - " but some logs (like bellow) generated automatically by Kamailio have no this information: Listening on > tcp: 172.20.0.4:5091 name tcp_socket advertise > 200.200.200.200:5091 > tls: 172.20.0.4:5063 name tls_socket advertise > 200.200.200.200:5063 > Aliases: > tls: 2eab5fb54462:5063 > tcp: 2eab5fb54462:5091 > 0(25) WARNING: tls [tls_init.c:807]: tls_h_mod_init_f(): openssl bug > #1491 (crash/mem leaks on low memory) workaround enabled (on low memory tls > operations will fail preemptively) with free memory thresholds 204800 and > 102400 bytes > 16(41) ERROR: tls [tls_server.c:1329]: tls_h_read_f(): protocol level error > 16(41) ERROR: tls [tls_util.h:42]: tls_err_ret(): TLS > accept:error:1408F09C:SSL routines:ssl3_get_record:http request > 16(41) ERROR: tls [tls_server.c:1333]: tls_h_read_f(): source IP: > 165.232.66.142 > 16(41) ERROR: tls [tls_server.c:1336]: tls_h_read_f(): destination IP: > 172.20.0.4 > 16(41) ERROR: <core> [core/tcp_read.c:1479]: tcp_read_req(): ERROR: > tcp_read_req: error reading - c: 0x7fc51458ef30 r: 0x7fc51458f058 (-1) -- Att, Benedito Marques

2 1

[Announcement] Philosophy of VoIP Infrastructure - An unusual free course
by Social Boh 13 Dec '25

13 Dec '25

Hi Kamailio community, I've published something unusual: a philosophy course for VoIP engineers. *"Philosophy of VoIP Infrastructure: Why Your Technical Decisions Are Philosophical"* Not "professional ethics 101." This is Plato, Foucault, Heidegger, and Bergson applied to actual Kamailio/RTPEngine problems. *Example:* The CAP theorem isn't just a technical limitation—it's the same debate Parmenides and Heraclitus had 2500 years ago about the nature of reality. Understanding this won't change your kamailio.cfg, but it will change how you think about consistency vs availability trade-offs. *Four modules:* - Ontology (what exists: protocols, containers, distributed state) - Epistemology (how we know: logs as testimony, Lamport timestamps, debugging biases) - Ethics (who benefits: power relations in centralized vs distributed architectures) - Temporality (lived vs measured time: jitter buffer as phenomenology) Each connects philosophical concepts with concrete VoIP code examples. It's completely free. Available in English and Spanish. Link: https://campus.mesaproyectos.com/course/view.php?id=53 *Warning:* This won't make you configure Kamailio faster. It will give you conceptual clarity that no RFC provides. If you're looking for a technical tutorial, this isn't it. If you've ever wondered why certain "technical" decisions feel philosophical, this course is for you. Feedback welcome. -- --- I'm SoCIaL, MayBe

2 2

Presence PUBLISH with a multiple branches
by Ihor Olkhovskyi 12 Dec '25

12 Dec '25

Hello, Got an issue and not sure what causes it. Architecture of a presence on the system is "edge" Kamailio is sending PUBLISH'es to "presence" Kamailio via "pua" and "pua_dialoginfo" modules and proxying SUBSCRIBE's from endpoints. "Presence" Kamailio basically just handle_subscribe()/handle_publish() with the config modparam("presence", "db_url", DBURL) modparam("presence", "subs_db_mode", 3) # DB-Only scheme. modparam("presence", "db_table_lock_type", 0) modparam("presence", "notifier_processes", 0) modparam("presence", "subs_htable_size", 15) modparam("presence", "delete_same_subs", 1) modparam("presence", "clean_period", 60) modparam("presence", "sip_uri_match", 1) modparam("presence", "pres_subs_mode", 0) modparam("presence", "timeout_rm_subs", 0) modparam("presence", "fetch_rows", 10000) modparam("presence", "pres_htable_size", 15) modparam("presence", "max_expires", 1800) modparam("presence", "send_fast_notify", 1) modparam("presence", "expires_offset", 10) On "edge" Kamailio several devices can be hidden behind same AOR, so on each branch of a call I'm getting PUBLISH with an "early" state, which is passed to "presence" Kamailio with local/remote tags and target-uri in XML. One of the devices are answering and hanging up, I'm getting PUBLISH with a state of a dialog (without from/remote tags, but target-uri). Which is completely fine. Problem is appearing on a call end, when a new SUBSCRIBE is coming to AOR, that was called with several devices behind. For whatever reason, "early" state of a device, that was locally CANCEL'ed by "edge" Kamailio is still there. As there were no "terminated" state generated by "edge" Kamailio for this particular target-uri? So, which Kamailio is need to be finetuned (and how?), one that is not sending PUBLISH for locally generated CANCEL or the one, that is not cleaning up dialog state completely? Thanks in advance! -- Best regards, Ihor

1 0

[uac] Data corruption in header restoration with long IMS URIs (RR-mode)
by Örn Arnarson 10 Dec '25

10 Dec '25

Hello everyone, I may have encountered a bug when replacing the from/to headers with UAC. This might be more for the sr-dev list, but I thought I'd try here first, in case my problem is caused by incorrect use. I am not sure whether or not it is related to the URIs being very long (as they are from an IMS), or some weirdness with the parameters. The issue is such; I receive a request from an internal IMS, that gets proxied to a third party after some header manipulation, such as from/to replacement. All provisional responses are handled correctly, but requests from the third party (such as re-INVITE or BYE) are failing with header restoration if using UAC in RR-mode (vsf and vst params). When using restore_dlg 1 modparam for uac, this does not happen. The error message is this: dialog [dlg_handlers.c:1343]: dlg_onroute(): unable to find dialog for BYE with route param 'ce.6a41' [236:5286] and call-id 'isbcldb8o4ztb7pwhh274d4zwb wubt2u7uwo(a)10.10.100.152' The Record-Route and Route headers are identical: Original INVITE Record-Route header: Record-Route: <sip:real.IP.address.removed;lr;ftag=ll8h2blu;did=ce.6a41;vsf=BwwcABwLBQUCCQIKW15TfR0IBEcBARcNFRoBTRQHEQBUTQYOGgxvbmFsO3Bob25lLWNvbnRleHQ9KzM1NA--;vst=AAAAABMABwENBwoORVxXLhZEE0EdHQgRGlNHBAAVGB0GQgxBRgINBjI3NC4zZ3BwbmV0d29yay5vcmdAaW1zLm1uYzAxMS5tY2MyNzQuM2dwcG5ldHdvcmsub3JnO3VzZXI9cGhvbmU-> And here's the Route header from the third party BYE request: Route: <sip:real.IP.address.removed;lr;ftag=ll8h2blu;did=ce.6a41;vsf=BwwcABwLBQUCCQIKW15TfR0IBEcBARcNFRoBTRQHEQBUTQYOGgxvbmFsO3Bob25lLWNvbnRleHQ9KzM1NA--;vst=AAAAABMABwENBwoORVxXLhZEE0EdHQgRGlNHBAAVGB0GQgxBRgINBjI3NC4zZ3BwbmV0d29yay5vcmdAaW1zLm1uYzAxMS5tY2MyNzQuM2dwcG5ldHdvcmsub3JnO3VzZXI9cGhvbmU-> Call-ID is identical as well. Decoding the BASE64 above seems to give some partial data, but I am not sure how it should look under "working" circumstances. The resulting From header is attempted to be restored, but is garbled: From: "TELNUM_REMOVED" <sip:TELNUM_REMOVED;phon'q$o,(??/bi50%#hu'~|6jbhW_PARTIAL_MCC_REMOVED_.3gppnetwork.org(a)ims.mnc_MNC_REMOVED_ .mcc_MCC_REMOVED_.3gppnetwork.org;user=phone>;tag=SD6sga699-04001233433931 Here's the original To (which is then the From trying to be restored in the BYE): To: <sip:telnum_removed;phone-context=ims.mncREMOVED.mccREMOVED.3gppnetwork.org@ims.mnc_MNC_REMOVED_ .mcc_MCC_REMOVED_.3gppnetwork.org;user=phone> Is this something anyone has encountered before? I am planning to use UAC mode, but I figured I should share this in case someone is unable to. Notable facts: kamailio 5.7.4, running on Ubuntu Server 24.04.3 LTS cfg_engine=python, using KSR. modparam("uac", "restore_mode", "auto") modparam("uac", "restore_dlg", 0) # to trigger the bug, 1 used in production to use dialog modparam("dialog", "dlg_flag", 4) KSR.dialog.dlg_manage() called before UAC replacement. NOTE: modparam("uac", "restore_dlg", 1) used for working scenario, as described above. I find it curious that the dialog module is complaining that it can't find the dialog, but yet it is doing some sort of partial replacement. Has anyone encountered something like this before? Thanks, Örn

2 2

Ability to dynamically add node to DMQ
by Rajneesh Soni 08 Dec '25

08 Dec '25

Hi, I have a Kamailio cluster with multiple nodes sharing information over DMQ. My DMQ configuration: - `notification_address` points to a DNS name that resolves to all cluster nodes - `multi_notify` is set to 1 **Problem:** When I add a new node to the cluster, I update the DNS record to include the new node's IP. However, DMQ only resolves the `notification_address` at startup, so existing nodes don't discover the new peer. **What I've tried:** 1. Looking for a DMQ reload command (like `dmq.reload`) - doesn't exist 2. `kamcmd cfg.reload` - doesn't re-resolve DMQ notification_address **Question:** How can I force DMQ to re-fetch/re-resolve the peer list without restarting Kamailio? Restarting is not an option as it would disrupt active TCP/WebSocket connections and ongoing calls. **Environment:** - Kamailio version: 6.0.1 - DMQ configuration: ``` modparam("dmq", "server_address",MY_DMQ_SERVER_ADDR) modparam("dmq", "notification_address", MY_DMQ_NOTIFICATION_ADDR) modparam("dmq", "multi_notify", 1) modparam("dmq", "num_workers", 4) modparam("dmq_usrloc", "enable", 1) modparam("dmq", "ping_interval", DMQ_PING_INTERVAL) ``` Any suggestions for dynamic peer discovery without restarts? Regards Rajneesh

3 4

2026

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 December 2025