November 2009 - FreeSWITCH-users - FreeSWITCH Mailing Lists

Holiday routing examples
by Andrew Thompson 30 Nov '09

30 Nov '09

Tony committed my patch for doing 'week of month' conditions in the XML dialplan along with some holiday routing examples to the default dialplan. Now you can detect all the major US holidays in pure dialplan XML without having to do any nasty math or anything (I did it all for you). I've also added a page to the wiki describing how to use it for other dates (like non-US holidays): http://wiki.freeswitch.org/wiki/Holiday_Routing Hope this helps some people. Andrew

2 1

Questions on ISDN support for Freeswitch
by Albano Daniele Salvatore - Lavoro 30 Nov '09

30 Nov '09

Hi to all, shortly i'll make a pbx for a customer that uses a couple of isdn bri lines and, looking for hardware, i've seen that not too much expensive isdn cards that works well are the ones that uses hfc-4/8s controller (specifically i'll use a OpenVox B200P that has 2 ISDN ports and use an HFC-4S controller). I've seen that FreeSwitch doesn't support mISDN but uses openzap (trough ozmod_isdn.so). I got some serious troubles using OpenZAP on analogical lines (bad dtmf recognition on fxs ports (like press 4 and get 44), annoying noise, busy/hangup tone unrecognized [tones, in italy, differs by cadency and not frequency], and more). Using tone_detect i bypassed the tone recognition problem and the noise was however acceptable: the blocking problem was the bad dtmf recognition. In the end (hope god forgive me) i put asterisk as (*1*)(*2*)sip proxy between zaptel and freeswich: i need to fix in config hangup cause detection but it seems to works fine. So my questions are: - Do freeswitch supports mISDN? - If it doesn't support mISDN, tone and dtmf recognition will be done by the isdn card, the kernel module or will be done by openzap? - There are alternative ways to use (*3*) mISDN with freeswitch, apart put asterisk as proxy? Thank for your support! Best Regards, Daniele --- (*1*) At beginning i tried using IAX but freeswitch segfaults when it try to answer the call and when i reload the module (trought reload mod_iax) shutdown routing didn't get called(the used iax library, smartly, start using another port without saying anything), however i need to do more testing: hope to open some tickets on jira in short. (*2*) I've seen that mod_iax config file support a "context" variable, but it isn't used so i wrote a small fix to use it if context isn't specified in the iax request (*3*) I need to use mISDN, rather other things, because i should use octasis soft echo cancellation and them supports only mISDN and Zaptel modules

1 0

Sangoma RTP TAP
by Helmut Kuper 30 Nov '09

30 Nov '09

Hello, has anyone of you tried the RTP TAP function of sangoma`s wanpipe driver? It is described here: http://wiki.sangoma.com/wanpipe-voice-rtp-tap On my side wanrouter log says that RTP TAB is configured and enabled, but I can't detect any udp packets received by the remote server (which is described by RTP_TAP_IP, RTP_TAP_MAC and RTP_TAP_PORT). I've latest driver and double checked the wanpipe.conf config. I tried to send some udp packets from wanpiping server to remote server, where the packets were shown up via tcpdump. So there is no FW problem involved. Each try to do some kind of printf debugging in wanpipe-driver doesn't succeed. Any ideas?

2 1

Call transfer fails in proxy media and bypass media modes in FreeSWITCH revision 15700
by John Platts 30 Nov '09

30 Nov '09

I have updated my FreeSWITCH installation to revision 15700. I am experiencing call transfer problems whenever proxy media or bypass media is enabled. When proxy media and bypass media are both disabled, the call transfer does not fail and there are no audio issues. When proxy media mode is enabled, the call stays up after the transfer occurs, but there is no audio flowing on either end of the call. When bypass media mode is enabled, there is no audio flowing on either end of the call, and the call actually gets disconnected. I have collected detailed traces using the TPORT_LOG=1 /usr/local/freeswitch/bin/freeswitch command. I have attached a ZIP file named freeswitch-rev15700-traces-112809-2210.zip, which includes the following traces: - freeswitch-rev15700-trace-112809-2210-proxyandbypassoff.txt - A trace with both media proxying and media bypass disabled. The call is being transferred without any problems in this scenario. - freeswitch-rev15700-trace-112809-2210-proxyonandbypassoff.txt - A trace with media proxying enabled and media bypass disabled. Media proxying is enabled for the call legs in this scenario. The call stays up in this scenario, but there is no audio flowing after the transfer completed. In this scenario, FreeSWITCH does not shutdown cleanly, and there is a segmentation violation when FreeSWITCH is terminated. - freeswitch-rev15700-trace-112809-2210-proxyandbypasson.txt - A trace with both media proxying and media bypass enabled. Media bypass is enabled for the call legs in this scenario. The call actually gets dropped and there is no audio after the transfer is completed in this scenario. I have looked over the SIP traces of the failing scenarios. I have caught the following problems in the failing scenarios: - The o= line in SDP descriptors coming from the IP phone contains the private IP address, but the c= line in the SDP descriptors coming from the IP phone contains the public IP address. I have noticed a problem in re-INVITEs being sent from in proxy media and bypass media modes. The c= line in the re-invites contains the private IP address instead of the public IP address. The c= line was modified by a SIP ALG to contain a public IP address, but FreeSWITCH is actually not handling this correctly when calls are transferred. - The wrong codec is being negotiated in re-INVITE to the transferred number in the scenario when media proxying is enabled but media bypass is disabled. - In the scenario where media bypass is used, the re-INVITE actually appears to contain the correct details, and we are receiving the correct responses from our IP to IP gateway, but FreeSWITCH is not handling the media streams properly. Example of SDP descriptor coming from IP phone (with SDP descriptor modified by SIP ALG): v=0 o=- 123576 123576 IN IP4 192.168.1.4 s=- c=IN IP4 173.57.44.212 t=0 0 m=audio 16406 RTP/AVP 18 0 8 2 9 104 101 a=rtpmap:18 G729/8000 a=rtpmap:0 PCMU/8000 a=rtpmap:8 PCMA/8000 a=rtpmap:2 G726-32/8000 a=rtpmap:9 G722/8000 a=rtpmap:104 L16/16000 a=rtpmap:101 telephone-event/8000 a=fmtp:101 0-15 a=ptime:20 a=sendrecv Notice that the c= line has the correct public IP address and the m= line containing the correct port. Example of incorrect SDP descriptor being sent by FreeSWITCH in re-INVITES: v=0 o=- 121397 121398 IN IP4 192.168.1.4 s=- c=IN IP4 192.168.1.4 t=0 0 m=audio 16404 RTP/AVP 18 0 8 101 a=rtpmap:18 G729/8000 a=fmtp:18 annexb=no a=rtpmap:0 PCMU/8000 a=rtpmap:8 PCMA/8000 a=rtpmap:101 telephone-event/8000 a=fmtp:101 0-15 a=sendonly a=ptime:20 Note that the c= line contains the wrong IP address, but the m= line contains the correct RTP port. Example of wrong re-INVITE message being sent to the number that the call was being transferred to: INVITE sip:19729831777@168.75.202.246:5060 SIP/2.0 Via: SIP/2.0/UDP 168.75.202.212:5062;rport;branch=z9hG4bKF1KrDreNFQgaj Max-Forwards: 69 From: "John Platts" <sip:19725357722@168.75.202.212>;tag=c61Drt38KF72m To: <sip:19729831777@ipipgw.ipdimensions.com>;tag=2B1339E0-1A2C Call-ID: 1c095553-5741-122d-33a8-00185167f91d CSeq: 123615824 INVITE Contact: <sip:mod_sofia@168.75.202.212:5062> User-Agent: FreeSWITCH-mod_sofia/1.0.trunk-15700M Allow: INVITE, ACK, BYE, CANCEL, OPTIONS, MESSAGE, UPDATE, INFO, REGISTER, REFER, NOTIFY Supported: timer, precondition, path, replaces Content-Type: application/sdp Content-Disposition: session Content-Length: 183 X-FS-Support: update_display Remote-Party-ID: "John Platts" <sip:19725357722@168.75.202.212>;party=calling;screen=yes;privacy=off v=0 o=- 123576 123577 IN IP4 192.168.1.4 s=- c=IN IP4 168.75.202.212 t=0 0 m=audio 30186 RTP/AVP 101 13 a=rtpmap:101 telephone-event/8000 a=fmtp:101 0-16 a=rtpmap:13 CN/8000 Here is the correct re-INVITE for the call that was unsuccessfully transferred (after the transfer was completed): INVITE sip:19729555871@168.75.202.246:5060 SIP/2.0 Via: SIP/2.0/UDP 168.75.202.212:5062;rport;branch=z9hG4bKgaDHFKZrc06vD Max-Forwards: 16 From: <sip:19725357722@168.75.202.212>;tag=BX8mpZj5p6ggS To: <sip:19729555871@168.75.202.246>;tag=2B12D184-BEC Call-ID: 15A1F95-DBD611DE-8C95D9DF-3419A306(a)168.75.202.246 CSeq: 123615820 INVITE Contact: <sip:19725357722@168.75.202.212:5062;transport=udp> User-Agent: FreeSWITCH-mod_sofia/1.0.trunk-15700M Allow: INVITE, ACK, BYE, CANCEL, OPTIONS, MESSAGE, UPDATE, INFO, REGISTER, REFER, NOTIFY Supported: timer, precondition, path, replaces Content-Type: application/sdp Content-Disposition: session Content-Length: 222 X-FS-Support: update_display v=0 o=- 121397 121399 IN IP4 192.168.1.4 s=- c=IN IP4 168.75.202.212 t=0 0 m=audio 26106 RTP/AVP 0 101 a=rtpmap:0 PCMU/8000 a=rtpmap:101 telephone-event/8000 a=fmtp:101 0-16 a=silenceSupp:off - - - - a=ptime:20 _________________________________________________________________ Windows 7: I wanted simpler, now it's simpler. I'm a rock star. http://www.microsoft.com/Windows/windows-7/default.aspx?h=myidea?ocid=PID24…

2 2

Polycom Phones and Domains
by Andrew Fritz 30 Nov '09

30 Nov '09

I'm attempting to configure several varieties of polycom (SoundPoint IP 550, SoundPoint IP 601) phones to connect to a freeswitch instance using a domain other than default (i.e. the ip address). Everything works wonderfully as long as the domain is named exactly the same thing as the server host provided to the phone (whether that is the server's ip address, or a hostname resolved via dns). As long as those match everything is fine. What I'm trying to sort out is, is it possible to convince the phone to use something other than the server's hostname/ip as the second part of the user name (i.e. user@host)? Or should I just resign myself to making the domain name some host name that can resolve via DNS? I've tried including @somedomain in the authid field of the line of the phone, but freeswitch reports that the user someuser@somedomain@serverip couldn't be authenticated. It appears that the phone always appends the server name... Does anyone know of a way that I configure a polycom to connect as user1@d1 and another to connect as user2@d2 where d1 and d2 are NOT in DNS or are polycoms just going to always use the server host as the @ part of the username? Andrew

1 0

Accessing custom SIP headers
by Simon Woodhead 30 Nov '09

30 Nov '09

Hi folks, I'm hoping someone can help me get at custom headers in the dial-plan. I've read about X- headers being accessible but need to get at some X_ headers passed through from a proxy. Reading the info app docs, the X shouldn't actually matter but no matter which way I try I always seem to get a null result. An example header in an INVITE is: X_ACCOUNTCODE: XXXXXX. I've tried the following dial-plan structures hoping one might work but none do: <action application="set" data="accountcodea=${variable_sip_h_X_ACCOUNTCODE}" /> <action application="set" data="accountcodeb=${variable_sip_h_ACCOUNTCODE}" /> <action application="set" data="accountcodec=${sip_h_X_ACCOUNTCODE}" /> <action application="set" data="accountcoded=${sip_h_ACCOUNTCODE}" /> Any help would be much appreciated. Thanks, Simon

2 1

CLIP on FXS channels with mod_openzap
by François Legal 30 Nov '09

30 Nov '09

Hello, I'm using Freeswitch with a Sangoma A400 card, and I'm having CLIP problems on the FXS ports. When I ring on FXS ports, the connected phone does not display callerid/callerid-name. I tried turning the stuff of in openzap.conf.xml () but it did not help. As a side note, turning this on on the FXO ports drops the callerid information on incoming calls. Running freeswitch 1.0.4 on linux 2.6.27. François

2 1

park on hook
by Woody Dickson 30 Nov '09

30 Nov '09

Hi, Is there anyway to detect when a channel is park in a way that is similar to hangup-hook or answer-hook? I would like to detect that inside a custom mod, without using the event mechanism? woody

1 0

DTMF Digits Lost when Under Load
by Michael Toop 30 Nov '09

30 Nov '09

Hi All, I have an issue that when my call volumes on my FS IVR box > 30 calls DTMF digits are lost (using RFC2833). It is definitely load related as it all works perfectly under 30 calls. Any pointers or a solution to the problem? Thanks, Michael

2 3

Callback to the user in ESL
by lakshmanan ganapathy 30 Nov '09

30 Nov '09

Hi, I'm using perl ESL to control the call in freeswitch. I'm having the following scenario, but not able to get it right. Dialplan: <extension name="outbound_soc"> <condition field="destination_number" expression="^9097$"> <action application="set" data="continue_on_fail=true"/> <action application="socket" data="192.168.1.222:8447 async full"/> </condition> </extension> 1. User A calls to an extention (1000). 2. My ESL program will be running, and it answers the call. 3. Then the program will get a number from the user. 4. It will hangup the call. 5. The program has to call to the number that was given by the user. In the above scenario, I was able to do until the 4th step. After hangup the call, if I say originate it is not working. Any ideas on how to do this in ESL.

4 8