sr-users May 2017

sr-users@lists.kamailio.org

91 participants
118 discussions

TLS SIP packet tracing and visualization
by Giovanni Maruzzelli 03 May '17

03 May '17

Hello fellows, after some experimentation with various tools, I come out with a little shell tool that maybe can be useful to you too. It can only work with non-forward secrecy ciphers, obviously, and only if is started before the client do the initial TLS handshake (eg, just restart the client). Forward secrecy cannot be decrypted after fact, so don't waste effort. An example of ciphers that can be decrypted are the "AES256-SHA" openssl cipher group. You can use ssldump to check what cipher is used by serverhello. Enjoy, make it better, and share it :) #!/bin/bash # brought to you by Giovanni Maruzzelli # SERVERIP="192.168.1.150" SERVERPORT="5061" PRIVKEY="/etc/certs/privkey.pem" STDERR2DEVNULL=" 2>/dev/null " REGEX="notyet" if [ -z "$1" ]; then REGEX="\\\.*" else REGEX="$1" fi FILTER="ssl.app_data and sip matches" FILTER2="$FILTER \"$REGEX\"" FILTER3="'$FILTER2'" ARGUMENT="-i 1 -Y $FILTER3 -E header=y -T fields -e frame.number -e frame.time -e frame.time_delta_displayed -e ip.src -e ip.dst -e sip.Status-Line -e sip.Request-Line -e sip.msg_hdr -l -d tcp.port\=\=5061,sip -o \"ssl.keys_list: $SERVERIP,$SERVERPORT,sip,$PRIVKEY\" $STDERR2DEVNULL | sed -u 's/\t/\n/g' | sed -u '/^$/d' | sed -u 's/^[0-9]*$/\n==&==============================/g'" echo "" echo "NB: if it do not works, edit script so that STDERR2DEVNULL=\" \" and try again" echo "" echo "NB: remember to quote and escape match patterns, using triple slash" echo " eg, for matching 1010(a)pbx.example.com, use \"1010(a)pbx.example.com \"" echo " eg, for matching anything, use \"\\\\\\.*\"" echo " eg, for matching *98, use \"\\\\\\*98\"" echo "USAGE: $0 \"\\\\\\*98(a)pbx.example.com\"" echo "" case "$1" in -help|--help|?) exit 0 ;; *) echo "THIS TIME WE'RE DOING:" echo "tshark $ARGUMENT" echo "" bash -c "tshark $ARGUMENT" ;; esac -- Sincerely, Giovanni Maruzzelli OpenTelecom.IT cell: +39 347 266 56 18

2 3

Memory leak when reloading htable using db_cluster
by Kristian F. Høgh 03 May '17

03 May '17

Hi list, I use current git master. When I reload a htable using "kamcmd htable.reload htable1", the "ctl handler" process leaks 384 bytes of pkg memory If I use a direct mysql connection without db_cluster, the is no memory leak modparam("db_cluster", "connection", "con1=>mysql://user:pass@ip.addr/database") modparam("db_cluster", "connection", "con2=>mysql://user:pass@ip.addr2/database") modparam("db_cluster", "cluster", "cls1=>con1=9s9s;con2=8s8s") modparam("htable", "db_url", "cluster://cls1") # modparam("htable", "db_url", "mysql://user:pass@ip.addr/database" modparam("htable", "htable", "htable1=>size=8;autoexpire=0;dbtable=htable1;") 16(10429) ALERT: qm_status: 6513. N address=0x7f4e65e01720 frag=0x7f4e65e016e8 size=128 used=1 16(10429) ALERT: qm_status: alloc'd from db_cluster: dbcl_api.c: db_cluster_init(294) 16(10429) ALERT: qm_status: start check=f0f0f0f0, end check= c0c0c0c0, abcdefed 16(10429) ALERT: qm_status: 6514. N address=0x7f4e65e01808 frag=0x7f4e65e017d0 size=128 used=1 16(10429) ALERT: qm_status: alloc'd from core: db.c: db_do_init2(298) 16(10429) ALERT: qm_status: start check=f0f0f0f0, end check= c0c0c0c0, abcdefed 16(10429) ALERT: qm_status: 6515. N address=0x7f4e65e018f0 frag=0x7f4e65e018b8 size=128 used=1 16(10429) ALERT: qm_status: alloc'd from core: db.c: db_do_init2(298) 16(10429) ALERT: qm_status: start check=f0f0f0f0, end check= c0c0c0c0, abcdefed Regards, Kristian Høgh Uni-tel A/S

2 8

SIP over WebSockets
by Nicolas Pace 02 May '17

02 May '17

Hi everyone, As I shared with you before, We are working on a mobile app to connect the Community Cellular Networks (that uses 2G Cellular Networks) with the rest of the world via SIP, and we are considering using Kamailio for this. We choosed to build a Progresive Web App (PWA) through WebSockets to reduce the friction involved in using a new tool. My question is, is there a reference config to use a web client through WebSockets? I don't have a clue on how to get from the stock config: https://gist.github.com/nicopace/9d44bf813fe309df69ff4cd0affe8324#file- _etc_kamailio_kamailio-conf to a fully functional WebSocket-enabled config. I've already read the WebSocket module documentation, but it is hard to understand as it assumes you already know how Kamailio works... too many moving parts :( I tried with https://github.com/caruizdiaz/kamailio-ws but it is giving me some errors: May 02 13:37:13 ubuntu kamailio[2329]: : <core> [cfg.y:3432]: yyerror_at(): parse error in config file /etc/kamailio/modules.cfg, line 48, column 12-25: failed to load module May 02 13:37:13 ubuntu kamailio[2329]: ERROR: <core> [cfg.y:3295]: yyparse(): cfg. parser: failed to find command ws_handle_handshake (params 0) May 02 13:37:13 ubuntu kamailio[2329]: : <core> [cfg.y:3435]: yyerror_at(): parse error in config file /etc/kamailio/kamailio.cfg, line 143, column 41: unknown command, missing loadmodule? Which relates with: https://github.com/caruizdiaz/kamailio-ws/blob/master/modules.cfg#L48 and https://github.com/caruizdiaz/kamailio-ws/blob/master/kamailio-ws.cfg#L 143 What I'm trying to accomplish is to have a PWA that sends SIMPLE messages and CALLS to and from SIP Users (the 2G users are behind a FreeSwitch, so it could be done). Just to give you some context, we will also add Push Notifications as 'ringing' or 'meesage recieved' alert, and that would be all (we also don't know how to do it). Well... hope to hear from you guys! Thanks!

2 2

TCP read/write buffer size on per connection or per module basis
by M S 02 May '17

02 May '17

Hi, I have a kamailio v4.4.5 deployment that serves TCP clients for SIP and XCAP requests (without SIP presence). The kamailio's core TCP parameters allow us to set TCP read and write buffer sizes which are set globally for all TCP connections. However, i noticed that in general SIP over TCP request is much smaller (<8KB for audio + video SIP INVITE request) then XCAP requests (say to support up to 500 contacts we need TCP buffers =~ 64KB). So, if I set global TCP buffer size according to XCAP service needs (e.g. 64KB), they waste too much RAM for SIP connections (assuming max SIP packet is 8KB, there is 56KB buffer space wasted per connection). On the other hand, if i set global TCP buffer sizes to SIP service needs (e.g. 8KB) then it is insufficient for XCAP (only 70-80 contacts would fix in XCAP list). Therefore, i am looking for a way to allow TCP buffer size settings per connection OR per module basis. Is it possible? If so then how? Thank you.

2 2

Retrieving the custom header parameter value from contact header
by vivek＠advaitamtech.com 02 May '17

02 May '17

Hi, Am trying to extract a custom parameter from the contact header. Below it the example of contact header, Eg: Contact: <sip:fayazat@XXX.XXX.XXX.XXX:5060;tgrp=tg_0615_prsw1> I want to extract the parameter value of "tgrp" Tried using @hf_value.contact.p.tgrp . But it is giving me complete contact header. I want only "tgrp" paramter value "tg_0615_prsw1". Anyone , any sugeestions? Thanks, Vivek.

3 3

TCP port re-using when doing HTTP transfer
by Ramachandran, Agalya (Contractor) 02 May '17

02 May '17

Hi Giacomo, I am using http_async_client module to do http transfer. Currently am testing the HTTP service with a mock server. If I send the HTTP response immediately for the POST request, then the TCP connections are reused from kamailio. In the mock server if I send the response with a delay of 1 sec then, TCP port are opened and closed for every HTTP transfer. Is there a way or any configuration to limit the number of TCP port opening for the HTTP transfers? Please clarify me. Regards, Agalya

2 1

Kamailio IMS - Handling 2 SIP transports in an IMS client
by Haggerty, Paul (phaggert) 01 May '17

01 May '17

Hi, Our client registers with the IMS core using UDP but uses TCP when the packet size is over 1300 bytes. When the client sends the TCP reques, the PCSCF replies with 403 Forbidden "you need to reg with the SCSCF" We've traced this back to the code in the PCSCF and it looks like it is checking the via headers and the received ip/port combinations for a match. The actual contact header looks like it matches the registered client. Question is what do we need to do in the client to ensure that the PCSCF recognizes the request as one from a registered client? We've tried the config options is_registered_fallback2ip , ignore_contact_rxport_check and they don't seem to be used. I'm attaching a network trace and included here are some debug messages from the PCSCF. In the network trace the client is 10.8.0.6. PCSCF Trace messages ------------------------------------------- 34207) DEBUG: ims_registrar_pcscf [service_routes.c:208]: getContactP(): This is a request - using first via to find contact 10(34207) DEBUG: ims_registrar_pcscf [service_routes.c:216]: getContactP(): searching for contact with host:port:proto contact [2://10.8.0.6:49946] 10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:457]: get_pcontact(): Searching for contact with AOR [sip:mcpttuser1@10.8.0.6:44580;ob;alias=10.8.0.6~44580~1;alias=10.8.0.6~49946~2] in P-CSCF usrloc based on VIA [2://10.8.0.6:49946] Received [2://10.8.0.6:49946], Search flag is 0 10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:463]: get_pcontact(): Have an AOR to search for 10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:468]: get_pcontact(): checking for rinstance10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:480]: get_pcontact(): no rinstance param 10(34207) DEBUG: ims_usrloc_pcscf [usrloc.c:148]: get_aor_hash(): Returning hash: [1662767777] 10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:503]: get_pcontact(): get_pcontact slot is [161] 10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:507]: get_pcontact(): comparing contact with aorhash [1662767777], aor [sip:mcpttuser1@10.8.0.6:44580;ob;alias=10.8.0.6~44580~1] 10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:508]: get_pcontact(): contact host [10.8.0.6:44580] 10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:509]: get_pcontact(): contact received [1:10.8.0.6:44580] 10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:543]: get_pcontact(): contact not found in memory 10(34207) INFO: ims_registrar_pcscf [service_routes.c:258]: getContactP(): Contact not found based on Contact-header, trying IP/Port/Proto 10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:457]: get_pcontact(): Searching for contact with AOR [sip:mcpttuser1@10.8.0.6:44580;ob;alias=10.8.0.6~44580~1;alias=10.8.0.6~49946~2] in P-CSCF usrloc based on VIA [2://10.8.0.6:49946] Received [2://10.8.0.6:49946], Search flag is 1 10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:463]: get_pcontact(): Have an AOR to search for 10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:468]: get_pcontact(): checking for rinstance10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:480]: get_pcontact(): no rinstance param 10(34207) DEBUG: ims_usrloc_pcscf [usrloc.c:148]: get_aor_hash(): Returning hash: [1662767777] 10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:503]: get_pcontact(): get_pcontact slot is [161] 10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:507]: get_pcontact(): comparing contact with aorhash [1662767777], aor [sip:mcpttuser1@10.8.0.6:44580;ob;alias=10.8.0.6~44580~1] 10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:508]: get_pcontact(): contact host [10.8.0.6:44580] 10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:509]: get_pcontact(): contact received [1:10.8.0.6:44580] 10(34207) DEBUG: ims_usrloc_pcscf [udomain.c:543]: get_pcontact(): contact not found in memory 10(34207) DEBUG: ims_registrar_pcscf [service_routes.c:263]: getContactP(): No entry in usrloc for 10.8.0.6:49946 (Proto 2) found! 10(34207) DEBUG: ims_registrar_pcscf [service_routes.c:300]: getContactP(): Asserted identity not set10(34207) ERROR: *** cfgtrace:request_route=[Orig_Initial] c=[/opt/OpenEPC/etc/pcscf2.cfg] l=975 a=26 n=xlog 10(34207) WARNING: <script>: 2.1 10(34207) ERROR: *** cfgtrace:request_route=[Orig_Initial] c=[/opt/OpenEPC/etc/pcscf2.cfg] l=976 a=26 n=send_reply Appreciate any help we can get, - Paul

1 0

handle_uri_alias persists through failure_route
by David Wilson 01 May '17

01 May '17

Hello All, We use registrar servers (call them A,B) for TLS connections from clients and a central routing proxy (P) that authenticates and manages all call routing. On INVITE, P uses lookup(“location”) and handle_uri_alias() to route the INVITE to A or B depending on the AOR from REGISTER. In failure_route I’m trying to send INVITE to a voicemail URI at host (V), using alias_db_lookup to retrieve it and sending with t_relay(). However, despite showing V as the host in the URI, the message is sent to A (or B) as if the handle_uri_alias is still in effect. I’ve tried calling revert_uri() before the alias_db_lookup(), but this doesn’t seem to help. Is there a way to reset the uri_alias? The failure_route (simplified) looks like this: failure_route[1] { if (method == "INVITE") { revert_uri(); if (alias_db_lookup("dbaliases")) { t_relay(); } } } I have also tried using t_relay(“$rd”, “$rp”) after confirming (by logging) that the uri contains the desired values, but kamailio failed to start, with this error logged: ERROR: tm [tm.c:672]: fixup_hostport2proxy(): TM module:fixup_hostport2proxy: bad port number <$rp> Any suggestions? Best regards, Dave.

3 3

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 May 2017