sr-users March 2005

sr-users@lists.kamailio.org

250 participants
627 discussions

by kcassidy＠kakelma.mine.nu

Hi All, I have a strange problem with my ser-0.8.14. I noticed that sometimes it's very slow in responding to requests (e.g. INVITES). After tinkering around, I saw my UDP queue gets filled up (netstat -an). The only fix I have right now is to restart the process. Then the problem will come up again around 1-2 days. I have around 40 registered users, and tried setting the children parameter in ser.cfg to different numbers like 4,6, and 10. But still having the problem. Any ideas? Thanks.

19 years, 8 months

[Serusers] 0.9.0 rpid bug?

by Daniel Poulsen

Perhaps this is a bug, perhaps it is by design? In ser 0.9.0, when i set debug=3, rpid ceases to function. Ethereal shows: [debug=3] Remote-Party-ID: 66570ab07228: authentication acc;party=calling;id-type=subscriber;privacy=none;screen=no [debug=2] Remote-Party-ID: sip:2125551212@sip.myhost.net;party=calling;id-type=subscriber;privacy=off;screen=no If this is a bug, let me know if I can provide more info. Dan

19 years, 8 months

[Serusers] rfc3261 and ACKs

by Steve Blair

Hello: Me again. I'm still stuck in the maze of call transfer and call forward all problems. Our gateway vendor suggested a new code release. Using this release completely breaks "normal" 5-digit and other calls that traverse the gateway. The vendor may be onto the problem but their assessment points back to my SER proxy. Given that the proxy works fine in all other calling scenarios with the old gateway code I'd like to know what is significant about their new code and SER. By the way the proxy in question is running 0.8.14 stable. The vendor problem description follows: Thanks,Steve --- cut here --- Here is what I believe the summary of the problem. I believe that the SIP Proxy is not complying with rfc3261 on how to process the ACK when Loose Routing is enabled. Debug Snippet from 12.3(13) ======================= *Mar 2 00:02:16.546: Received: SIP/2.0 200 OK Via: SIP/2.0/UDP 128.91.56.38:5060 From: <sip:343408@128.91.56.38>;tag=5286C18-13B5 To: <sip:68001@net.isc.upenn.edu>;tag=003094c38fd4003a29fa1f0f-264a55c1 Call-ID: 2C5BA6F-15B811CC-8122ED93-3017C386(a)128.91.56.38 Date: Thu, 17 Mar 2005 12:08:16 GMT CSeq: 101 INVITE Server: CSCO/7 Contact: <sip:68001@128.91.56.6:5060> Record-Route: <sip:68001@128.91.56.47;ftag=5286C18-13B5;lr=on> *Mar 2 00:02:16.578: Sent: ACK sip:68001@128.91.56.47:5060;ftag=5286C18-13B5;lr=on SIP/2.0 Via: SIP/2.0/UDP 128.91.56.38:5060 From: <sip:343408@128.91.56.38>;tag=5286C18-13B5 To: <sip:68001@net.isc.upenn.edu>;tag=003094c38fd4003a29fa1f0f-264a55c1 Date: Tue, 02 Mar 1993 00:02:15 GMT Call-ID: 2C5BA6F-15B811CC-8122ED93-3017C386(a)128.91.56.38 Route: <sip:68001@128.91.56.6:5060> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ The IP Phones address and that is why the call proceeds further. Because the ACK gets to the phone Debug Snippet from 12.3(11)T3 ========================= As of this code release, the gateways added RFC3261 support for "Loose Routing". Received: SIP/2.0 200 OK Via: SIP/2.0/UDP 128.91.56.38:5060;branch=z9hG4bK0F86 From: <sip:4843433408@128.91.56.38>;tag=60040-250 To: <sip:68001@net.isc.upenn.edu>;tag=003094c38fd4003c61e999c2-43d890a2 Call-ID: 4479EFAB-2BF611D6-800BA01F-9B2E2BA3(a)128.91.56.38 Date: Thu, 17 Mar 2005 12:30:36 GMT CSeq: 101 INVITE Server: CSCO/7 Contact: <sip:68001@128.91.56.6:5060> Record-Route: <sip:68001@128.91.56.47;ftag=60040-250;lr=on> *Mar 1 02:52:58.151: //-1/xxxxxxxxxxxx/SIP/Msg/ccsipDisplayMsg: Sent: ACK sip:68001@128.91.56.6:5060 SIP/2.0 Via: SIP/2.0/UDP 128.91.56.38:5060;branch=z9hG4bK116B8 From: <sip:4843433408@128.91.56.38>;tag=60040-250 To: <sip:68001@net.isc.upenn.edu>;tag=003094c38fd4003c61e999c2-43d890a2 Date: Fri, 01 Mar 2002 02:52:57 GMT Call-ID: 4479EFAB-2BF611D6-800BA01F-9B2E2BA3(a)128.91.56.38 Route: <sip:68001@128.91.56.47;ftag=60040-250;lr=on> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Address of the Loose Router which is compliant to route processing in RFC3261 The SIP Proxy (Which is the loose router) should receive this ACK and forward onto the IP Phone or the address in the Req-URI which it isn't. The proxy is actually acting like a Strict router in this instance. The Proxy is adding the loose routing parameters in the headers but doesn't seem to processing the Requests correctly as a loose router. So that is why we are having issues with a basic call in this code release.

19 years, 8 months

[Serusers] I need serial forking example

by PHAN, Quang-Minh (Docteur Ordinateur)

Hi List, Can anyone send me an example of serial forking using avpops module. I remember reading one tutorial somewhere but I couldn't find it again. Thanks for your help Minh

19 years, 8 months

[Serusers] Problems inspecting SER with ngrep

by Alberto Cruz

Hi all, I'm trying to inspect the packets that are arriving to my SER server with ngrep but when I executed the command I got the following: [root@matrix root]# ngrep @sip.telereunion.com.mx port 5060 interface: eth0 (65.208.39.208/255.255.255.240) filter: ip and ( port 5060 ) match: @sip.telereunion.com.mx ## As you can see I'm only getting hash symbol. Does anybody know how could I display the right info? I'm running RedHat Linux 9.0, ser 0.9.1, ngrep 1.43 and libcap 0.8.3 Regards Alberto Cruz

19 years, 8 months

[Serusers] additional contact info from grandstream budgetone ?

by maka

Hi all, i just got a gradnstream budgetone for testing, and I see some strange behaviour on its part - it 's adding an additional ;user=phone string to the contact field it sends to the registrar, and it thus produces some bad header errors when communicating with the phone. the phone also gets registered twice for the same number -- once with the extra information from above, and once without it. has anyone dealt with something like this?

19 years, 8 months

[Serusers] Syslog options

by Marc Khayat

hi all, i'm using ser-0.8.14, and i'm receiving call logs on a syslog server. But i'm not getting the call duration: in the below logs, SER is linux.domain.net, 192.168.0.150 is registered, 65.25.48.112 is not registered but stateful routing is used (t_relay_to_udp) Mar 17 17:22:56 linux /sbin/ser[10908]: ACC: transaction answered: method=INVITE, i-uri=sip:33@linux.domain.net, o-uri=sip:33@linux.domain.net, call_id=0318F319-981E-4AD7-86E8-1DA55DB661C3(a)192.168.0.150, from=Marc Khayat <sip:777@linux.domain.net:5061>;tag=3813796527, code=200, to=<sip:33@linux.domain.net>;tag=b5422284a4 Mar 17 17:22:56 linux /sbin/ser[10912]: ACC: request acknowledged: method=ACK, i-uri=sip:33@65.25.48.112, o-uri=sip:33@65.25.48.112, call_id=0318F319-981E-4AD7-86E8-1DA55DB661C3(a)192.168.0.150, from=Marc Khayat <sip:777@linux.domain.net:5061>;tag=3813796527, code=200, to=<sip:33@linux.domain.net>;tag=b5422284a4 Mar 17 17:22:58 linux /sbin/ser[10915]: ACC: transaction answered: method=BYE, i-uri=sip:33@65.25.48.112, o-uri=sip:33@65.25.48.112, call_id=0318F319-981E-4AD7-86E8-1DA55DB661C3(a)192.168.0.150, from=Marc Khayat <sip:777@linux.domain.net:5061>;tag=3813796527, code=200, to=<sip:33@linux.domain.net>;tag=b5422284a4 I tried the options in log_fmt, but there's nothing. How is it best to do accounting if syslog will not give me the call duration? would there be any difference (call duration, ASR, ...) if I use a Radius or MySQL ? Thanks!

19 years, 8 months

RE: [Serusers] Mediaproxy consuming RAM resources.

by Ricardo Martinez

Paul. Hello. Sorry to interrupt you. I'm still stuck with the mediaproxy resource problem. I've installed the patch from the pyhton.org but the problem seems to persist. I'm thinking maybe the problem is related with the way i treat a NAT'd call. I hope that you can help me with some questions. I have a mix NAT'd solution, the mayority of my endpoint have STUN enabled, so if this first solution fails, the mediaproxy is enabled for those calls. Regarding to the mediaproxy : What i do is to check if the caller is NAT'd if so, the mediaproxy is enabled. When a endpoint with a public ip call to a NAT'd endpoint i do the same check but in the on_reply_route. Is this ok?. How you handle a call when a endpoint in the public internet call to another in the public internet too? Do you use mediaproxy in that case?. I'm attaching my ser.cfg to check if there is any problem with it. I really appreciate any help Regards, Ricardo.- > -----Mensaje original----- > De: Java Rockx [mailto:javarockx@gmail.com] > Enviado el: Miércoles, 16 de Marzo de 2005 10:22 > Para: Ricardo Martinez > CC: serusers(a)lists.iptel.org > Asunto: Re: [Serusers] Mediaproxy consuming RAM resources. > > > That's what I have. I'm not sure what the setflag(1) is for, but then > again, I cannot see the entire ser.cfg. > > Regards, > Paul > > > On Wed, 16 Mar 2005 10:20:41 -0400, Ricardo Martinez > <rmartinez(a)redvoiss.net> wrote: > > Hello Paul. > > I have the next lines for ending media session: > > > > if (loose_route()) { > > if (method=="BYE" || method=="CANCEL") { > > end_media_session(); > > setflag(1); > > }; > > t_relay(); > > break; > > }; > > > > Is this ok?. I got the idea from a ser.cfg file in the > forum. I'm not so > > sure if this is correct. > > Can someone help me here? > > > > Thanks again! > > > > Regards, > > Ricardo > > > > > -----Mensaje original----- > > > De: Java Rockx [mailto:javarockx@gmail.com] > > > Enviado el: Martes, 15 de Marzo de 2005 21:03 > > > Para: serusers(a)lists.iptel.org > > > Asunto: Re: [Serusers] Mediaproxy consuming RAM resources. > > > > > > > > > Yes that could help, but if you're experiencing pleny of > "idle" RTP > > > connections then you are probably not ending your media > sessions with > > > end_media_session(). > > > > > > Regards, > > > Paul > > > > > > > > > On Tue, 15 Mar 2005 18:33:19 -0600, Alberto Cruz > > > <acruz(a)tekbrain.com> wrote: > > > > If you are running RedHat 9 and Python 3.2.5 may be this > > > could help you: > > > > > > > > http://www.python.org/download/download_linux.html > > > > > > > > Regards > > > > > > > > Alberto Cruz > > > > > > > > > > > > Ricardo Martinez wrote: > > > > > > > > Hello. Returning to this subject. I upgraded to the new > > > mediaproxy version. > > > > 1.2.1, but i still have the same problem. I detected a > > > increasing amount of > > > > RAM memory used by mediaproxy. I don't know what's > > > happening. Is this > > > > something maybe related to the way i tear dow the session, > > > i'm still confuse > > > > with so many IDLE session in my mediaproxy. Can this be the > > > cause of the > > > > increasing amount of RAM memory used? Also, is there a way > > > to debug the > > > > mediaproxy? or something that help me to understand what's > > > going on? Any > > > > help would be appreciated. Thanks Ricardo.- > > > > -----Mensaje original----- De: Java Rockx > > > [mailto:javarockx@gmail.com] > > > > Enviado el: Viernes, 11 de Marzo de 2005 15:23 Para: > > > Ricardo Martinez CC: > > > > serusers(a)lists.iptel.org Asunto: Re: [Serusers] Mediaproxy > consuming RAM > > > > resources. I'm tearing down my sessions the same way you > > > showed in your > > > > first email. So AFAIK, that is OK. Regards, Paul On > Fri, 11 Mar 2005 > > > > 16:15:38 -0400, Ricardo Martinez <rmartinez(a)redvoiss.net> wrote: > > > > Paul, I was thinking the same, maybe that's the problem. > > > What about the > > > > tear down of the sessions? Thanks! Regards, Ricardo.- > > > > -----Mensaje original----- De: Java Rockx > > > [mailto:javarockx@gmail.com] > > > > Enviado el: Viernes, 11 de Marzo de 2005 15:09 Para: > > > Ricardo Martinez CC: > > > > serusers(a)lists.iptel.org Asunto: Re: [Serusers] Mediaproxy > consuming RAM > > > > resources. Ricardo, I think the release notes on > > > mediaproxy-1.2 say that > > > > some resource leaks have been fixed. Perhaps you can > try version 1.2 > > > > Regards, Paul On Fri, 11 Mar 2005 16:01:59 -0400, > Ricardo Martinez > > > > <rmartinez(a)redvoiss.net> wrote: > > > > Hello list. I was looking the process in my SER > sip-proxy with the > > > > mediaproxy running. > > > > Every time i look, the percetage of used RAM is > > > increasing. It seems like > > > > for some reason mediaproxy is not releasing the resources > > > used in RAM. You > > > > can look this in the next lines : PID USER PRI NI SIZE RSS > > > SHARE STAT %CPU > > > > %MEM TIME CPU COMMAND > > > > 1952 root 15 0 284M 266M 1628 S 0.0 26.4 19:22 0 > > > > mediaproxy.py Can someone tell me if this is a bug of > > > mediaproxy or maybe a > > > > missconfiguration in my ser.cfg? I want to ask a question > > > regarding to tear > > > > down sessions with mediaproxy > > > > also. In my ser.cfg i have : if (loose_route()) { if > > > (method=="BYE" || > > > > method=="CANCEL") { log(1, "NAT: BYE o CANCEL recibido --> > > > terminando la > > > > sesion de media\n"); end_media_session(); setflag(1); }; > > > t_relay(); break; > > > > }; Is this ok? I think in some cases (i'm still unable to > > > determine which > > > > ones), SER does not tear down a mediaproxy session. You > > > can see for example > > > > this two session, both of them were ended, but it seems to > > > be active even > > > > before that. (i ommited the source, via and destination > > > IP's in the output > > > > of the session.py) Status Duration Codec Type Traffic > > > > -------------------------------------------------------------- > > > > -------------- > > > > --- hold (2430) 40'40" G729 Audio 11.53k/29.58k/18.45k > > > inactive 0'02" > > > > Unknown Audio 0/0/0 Can someone tell me what's happening? > > > I'm runnig : [root > > > > mediaproxy]# ./mediaproxy.py --version mediaproxy.py 1.0 > > > Thanks in advance. > > > > Regards, Ricardo M. > > > _______________________________________________ Serusers > > > > mailing list serusers(a)lists.iptel.org > > > > http://lists.iptel.org/mailman/listinfo/serusers > > > > _______________________________________________ Serusers > > > mailing list > > > > serusers(a)lists.iptel.org > http://lists.iptel.org/mailman/listinfo/serusers > > > > _______________________________________________ > > > > Serusers mailing list > > > > serusers(a)lists.iptel.org > > > > http://lists.iptel.org/mailman/listinfo/serusers > > > > > > > > > > > > > > > > > > _______________________________________________ > > > Serusers mailing list > > > serusers(a)lists.iptel.org > > > http://lists.iptel.org/mailman/listinfo/serusers > > > > > >

19 years, 8 months

[Serusers] just noticed a strange problem with aliases...they dont work instantly

by Iqbal

Hi If I add a alias manually into my DB, they dont seem to go live instantly, serctl, doesnt show em...I guess thats because its reading from memory, if this is the case, then how can I get the entries in the DB into memory...without restarting. Or shoudk I not add directly into the DB, and go via serctl...which is not really want I want to do... Iqbal

19 years, 8 months

RE: [Serusers] LCR gw config problem?

by Matt Schulte

smallint doesn't seem to allow null, when I change to varchar to allow null it breaks the whole deal... -----Original Message----- From: Juha Heinanen [mailto:jh@lohi.tutpro.com] Sent: Thursday, March 17, 2005 8:47 AM To: Matt Schulte Cc: serusers(a)lists.iptel.org Subject: [Serusers] LCR gw config problem? Matt Schulte writes: > I'll keep it short, when trying to use the gw lookup everything works > fine except the port gets altered to some crazy random port number. > Thoughts? hard to say. it works for me. if you use port 5060, you can give port field NULL value. -- juha

19 years, 8 months

← Newer
1
...
27
28
29
30
31
32
33
...
63
Older →

Jump to page:

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 March 2005