sr-users September 2010

sr-users@lists.kamailio.org

58 participants
97 discussions

mi ul_dump 'buffer too small'
by Juha Heinanen 22 Sep '10

22 Sep '10

is there anything one can do when 'mi ul_dump' gives error: 500 - Internal server error processing 's': buffer too small i guess there is too many contacts. -- juha

3 5

Modifying $pU
by dotnetdub 22 Sep '10

22 Sep '10

Hi All I am receiving a Preferred-Identity from a customer which I need to modify (strip the first character a 0) and then send it as Remote Party ID Sending the RPID is no problem, my issue is stripping the character. I have tried a few things to do this. subst('/^P-Preferred-Identity:(.*)<sip:(0)(.*)@.*>(.*)$/P-Preferred-Identity:\1<sip:\3@domain >\4/ig') This substitution works fine but when I read $pU it still has the leading 0 I also looked at using re.subst on $pU but can't get it to work. What is the best way to achieve what I'm trying to do? Sorry for hassling you! Thanks, Brian

5 5

[OT] Any utility to create cool SIP flows in HTML format?
by Iñaki Baz Castillo 21 Sep '10

21 Sep '10

Hi, I'm trying to create cool SIP flows to be displayed in a HTML. What I'm looking for is something like (or exactly) this: http://www.tech-invite.com/Ti-sip-dialog.html There you can click into messages to show the complete headers/body and so on. Does somebody know a tool to create such flows? Thanks for any suggestion. NOTE: There are some others utilities that allow creating TXT flows, as in a draft/RFC. I'm looking for something really more powerful than that. -- Iñaki Baz Castillo <ibc(a)aliax.net>

3 5

Change in RTP Payload time on the fly
by Pranav Desai 20 Sep '10

20 Sep '10

Hello, We have a proxy server setup that runs Kamailiov3.0.1 + RTPProxyv1.2.1. Nathelper support has been enabled in Kamailio. On a certain route, we call force_rtp_proxy("z120"), i.e. with resizing parameter 'z' and argument 120ms. During a call where endpoints have negotiated the use of G729, RTPProxy sends packets with ptime=12 (120ms), however occasionally rtpproxy sends packets with ptime 6(60ms), 8(80ms), 4(40ms). Under what circumstance would RTPProxy send packets with ptime < 12 ? Regards, Pranav Desai

4 5

Request for consulting assistance with SER scripts and setup
by Scott Seltzer 20 Sep '10

20 Sep '10

If there is anyone who would be able to provide assistance with SER programming, configuration changes, or migrations to SER3.0 please send email to: Scott(a)connectmevoice.com Thanks , Scott Seltzer

1 0

Re: [SR-Users] Usrloc reload problem
by Jan Klepal 19 Sep '10

19 Sep '10

Hello Daniel, On Saturday September 18 2010 12:03:59 Daniel-Constantin Mierla wrote: > Hello, > > On 9/17/10 3:27 PM, Jan Klepal wrote: > > Hello, > > > > I run kamailio 3.0.3 with location table stored in database (PostgreSQL > > 8.4.2) with db_mode 2 (write-back) and upon kamailio restart some > > contacts from database are loaded incorrectly. Messages for those > > contacts are routed to invalid IP addresses. > > > > After some debugging I figured out that it happens only for contacts > > which "received" column value is null. For those contacts value is > > filled from previously loaded contact having non-null value in > > "received" column. > > > > it was a bug, hopefully fixed. Inside the postgres module, an internal > buffer was not reset. > > Test with master branch or take the patch from commit (it is small) and > do it with 3.0.x: > http://lists.sip-router.org/pipermail/sr-dev/2010-September/008994.html > > Since I have no pg server at hand, couldn't actually test it, so when > you report is working fine, I will backport to 3.0.x. > > Thanks, > Daniel Thank you for fast reply. Yes, this patch solved the problem. Contacts are loaded correctly from DB now. Kind Regards, Jan

2 1

Re: [SR-Users] Change in RTP Payload time on the fly
by Pranav Desai 19 Sep '10

19 Sep '10

Hello, Thank you for your comments. My answers are inline. > Hello, > > ptime and also the z flag are ms values, so a values of 6 or 8 ms are not > the recommended length of time, it should be an integer > multiple of 5 ms. I havn't yet a look the the source, but the nathelper > doc > say ZNN, maybe only the 12 is taken from your 120? The 'ZNN' in the documentation might be misleading, since I tried with 'Z120' and the variable 'requested_nsamples' is set to 960 samples (120/10*80). This variable is used by RTP resizer to resize the RTP packets. > But in fact the doc confuse with ZNN and an example of 100ms. Could you > try > again with 90? > I tried with ptime=80ms and ptime=60ms. But, I still have the similar results, i.e. I get occasional packets with ptime=40 and sometimes ptime=20ms. Regards, Pranav Desai > regards, > Andreas > > 2010/9/18 Pranav Desai <pdesai(a)signalogic.com> > >> >> Hello, >> >> We have a proxy server setup that runs Kamailiov3.0.1 + RTPProxyv1.2.1. >> Nathelper support has been enabled in Kamailio. >> >> On a certain route, we call force_rtp_proxy("z120"), i.e. with resizing >> parameter 'z' and argument 120ms. During a call where endpoints have >> negotiated the use of G729, RTPProxy sends packets with ptime=12 >> (120ms), >> however occasionally rtpproxy sends packets with ptime 6(60ms), 8(80ms), >> 4(40ms). >> >> Under what circumstance would RTPProxy send packets with ptime < 12 ? >> >> Regards, >> Pranav Desai >> >> _______________________________________________ >> SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing list >> sr-users(a)lists.sip-router.org >> http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users >> >

1 0

500 command not activated
by dotnetdub 18 Sep '10

18 Sep '10

Hi All, Kamailio 3.0.3 kamctl domain show 500 command not activated however I can run kamctl domain showdb which dumps out the domains from the database. What am I missing? Thanks Brian

2 5

Usrloc reload problem
by Jan Klepal 18 Sep '10

18 Sep '10

Hello, I run kamailio 3.0.3 with location table stored in database (PostgreSQL 8.4.2) with db_mode 2 (write-back) and upon kamailio restart some contacts from database are loaded incorrectly. Messages for those contacts are routed to invalid IP addresses. After some debugging I figured out that it happens only for contacts which "received" column value is null. For those contacts value is filled from previously loaded contact having non-null value in "received" column. Here is sample debug, which shows contact (row #97 in result set) with IP 10.3.71.1 is later used for contact without "received" value (row #172): db_postgres [km_dbase.c:149]: 0x888e00 PQsendQuery(select username,contact,expires,q,callid,cseq,flags,cflags,user_agent,received,path,socket,methods,last_modified from location ) ... db_postgres [km_res.c:242]: PQgetvalue(0x888e00,79,9)=[sip:10.3.71.1:5060] db_postgres [km_res.c:251]: [79][9] Column[received]=[sip:10.3.71.1:5060] ... db_postgres [km_res.c:242]: PQgetvalue(0x888e00,172,9)=[] ... db_postgres [km_res.c:242]: PQgetvalue(0x888e00,172,0)=[222333444] db_postgres [km_res.c:251]: [172][0] Column[username]=[222333444] db_postgres [km_res.c:242]: PQgetvalue(0x888e00,172,1)=[sip:222333444@195.70.1.200] db_postgres [km_res.c:251]: [172][1] Column[contact]=[sip:222333444@195.70.1.200] db_postgres [km_res.c:242]: PQgetvalue(0x888e00,172,2)=[2010-09-16 00:21:04] db_postgres [km_res.c:251]: [172][2] Column[expires]=[2010-09-16 00:21:04] db_postgres [km_res.c:242]: PQgetvalue(0x888e00,172,3)=[-1] db_postgres [km_res.c:251]: [172][3] Column[q]=[-1] db_postgres [km_res.c:242]: PQgetvalue(0x888e00,172,4)=[0fc6692b6570aaac04b3c85a3b9496c0(a)sipxsip.cz] db_postgres [km_res.c:251]: [172][4] Column[callid]=[0fc6692b6570aaac04b3c85a3b9496c0(a)sipxsip.cz] db_postgres [km_res.c:242]: PQgetvalue(0x888e00,172,5)=[397] db_postgres [km_res.c:251]: [172][5] Column[cseq]=[397] db_postgres [km_res.c:242]: PQgetvalue(0x888e00,172,6)=[0] db_postgres [km_res.c:251]: [172][6] Column[flags]=[0] db_postgres [km_res.c:242]: PQgetvalue(0x888e00,172,7)=[0] db_postgres [km_res.c:251]: [172][7] Column[cflags]=[0] db_postgres [km_res.c:242]: PQgetvalue(0x888e00,172,8)=[Asterisk SIPXSIP] db_postgres [km_res.c:251]: [172][8] Column[user_agent]=[Asterisk SIPXSIP] db_postgres [km_res.c:242]: PQgetvalue(0x888e00,172,9)=[] db_postgres [km_res.c:242]: PQgetvalue(0x888e00,172,10)=[] db_postgres [km_res.c:242]: PQgetvalue(0x888e00,172,11)=[udp:80.200.1.100:5060] db_postgres [km_res.c:251]: [172][11] Column[socket]=[udp:80.200.1.100:5060] db_postgres [km_res.c:242]: PQgetvalue(0x888e00,172,12)=[] db_postgres [km_res.c:242]: PQgetvalue(0x888e00,172,13)=[2010-09-16 00:11:04] db_postgres [km_res.c:251]: [172][13] Column[last_modified]=[2010-09-16 00:11:04] <core> [db_row.c:119]: allocate 448 bytes for row values at 0x998120 <core> [db_val.c:117]: converting STRING [222333444] <core> [db_val.c:127]: allocate 10 bytes memory for STRING at 0x944008 <core> [db_val.c:117]: converting STRING [sip:222333444@195.70.1.200] <core> [db_val.c:127]: allocate 27 bytes memory for STRING at 0x9378d8 <core> [db_val.c:157]: converting DATETIME [2010-09-16 00:21:04] <core> [db_val.c:106]: converting DOUBLE [-1] <core> [db_val.c:117]: converting STRING [0fc6692b6570aaac04b3c85a3b9496c0(a)sipxsip.cz] <core> [db_val.c:127]: allocate 44 bytes memory for STRING at 0x925cd8 <core> [db_val.c:73]: converting INT [397] <core> [db_val.c:73]: converting INT [0] <core> [db_val.c:73]: converting INT [0] <core> [db_val.c:117]: converting STRING [Asterisk SIPXSIP] <core> [db_val.c:127]: allocate 17 bytes memory for STRING at 0x9061e8 <core> [db_val.c:117]: converting STRING [sip:10.3.71.1:5060] <core> [db_val.c:127]: allocate 19 bytes memory for STRING at 0x905fc8 <core> [db_val.c:56]: converting NULL value <core> [db_val.c:117]: converting STRING [udp:80.200.1.100:5060] <core> [db_val.c:127]: allocate 22 bytes memory for STRING at 0x905da8 <core> [db_val.c:73]: converting INT [4767] <core> [db_val.c:157]: converting DATETIME [2010-09-16 00:11:04] To make things clear, all rows after row #97 had "received" value null, and row #172 was UA I ran tests on... in the other words contact with null "received" value is filled with previous non-null value as long next non-null row is fetched from DB and following null contacts are filled with its value and so on. When the contact having invalid "registered" re-register it will overwrite value in memory and calls are routed to proper IP address. This query is sent to database then: db_postgres [km_dbase.c:149]: 0x888e00 PQsendQuery(update location set expires='2010-09-16 00:30:58',q=-1.00 ,cseq=103,flags=0,cflags=0,user_agent='Asterisk SIPXSIP',received=NULL,path=NULL,socket='udp:80.200.1.100:5 060',methods=NULL,last_modified='2010-09-16 00:20:58' where username='222333444' AND contact='sip:222333444@195.70.1.200' AND callid='0fc6692b6570aaac04b3c85a3b9496c0(a)sipxsip.cz') I've also checked status of this record in usrloc by "kamctl ul show"... both before and after kamailio restart state was CS_SYNC. Is this bug or should I somehow force kamailio to store "received" into database? Kind Regards, Jan Klepal

2 1

Kamailio Jabberd Integration
by shoks 17 Sep '10

17 Sep '10

Hi, Has anyone managed to integrate Kamailio and Jabberd to support presence and messaging functions (XMPP). I've tried to setup based on notes and emails that I found online, but still could'nt make it work. I appreciate if someone can share their configuration or give me a link. Thanks, sho

3 4

← Newer
1
2
3
4
5
6
7
8
9
10
Older →

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 September 2010