March 2009 - freeswitch-dev@lists.freeswitch.org

[Freeswitch-dev] Make waiting around an option for consumer callbacks in mod_fifo [patch]
by Herman Griffin 16 Mar '09

16 Mar '09

The consumer callback that has been added to mod_fifo is great. I've been looking for this feature for a long time. However, in it's current state it still has the problem that I was trying to avoid in the first place; I don't want agents to just be sitting around waiting for callers with their phone off the hook. I made a slight modification mod_fifo.c that makes the "waiting around" an option. If you set the channel following channel variable in the originate string for agents in fifo.conf.xml, then the consumers phone will either hangup or stay off hook when they answer a queue calls: member_wait = wait|nowait nowait: agents phone hangs up wait: agents wait around I'm including a patched for mod_fifo.c and fifo.conf.xml. li53-164:~/fstrunk# svn diff src/mod/applications/mod_fifo/mod_fifo.c conf/autoload_configs/fifo.conf.xml Index: src/mod/applications/mod_fifo/mod_fifo.c =================================================================== --- src/mod/applications/mod_fifo/mod_fifo.c (revision 12607) +++ src/mod/applications/mod_fifo/mod_fifo.c (working copy) @@ -462,10 +462,23 @@ channel = switch_core_session_get_channel(session); + + /*Add option to wait on not wait*/ + const char *member_wait = NULL; + member_wait = switch_channel_get_variable(channel, "member_wait"); + + /*if the member_wait channel variable is NULL or not set to "wait" or 'nowait" + then default to "wait" + */ + if( NULL == member_wait || !( !strcasecmp(member_wait, "wait") || !strcasecmp(member_wait, "nowait") ) ) + { + member_wait = "wait"; + } + switch_channel_set_variable(channel, "fifo_outbound_uuid", h->uuid); switch_core_event_hook_add_state_change(session, hanguphook); app_name = "fifo"; - arg = switch_core_session_sprintf(session, "%s out wait", h->node_name); + arg = switch_core_session_sprintf(session, "%s out %s", h->node_name, member_wait); /*Add option to wait or not wait*/ extension = switch_caller_extension_new(session, app_name, arg); switch_caller_extension_add_application(session, extension, app_name, arg); switch_channel_set_caller_extension(channel, extension); Index: conf/autoload_configs/fifo.conf.xml =================================================================== --- conf/autoload_configs/fifo.conf.xml (revision 12607) +++ conf/autoload_configs/fifo.conf.xml (working copy) @@ -1,7 +1,7 @@ <configuration name="fifo.conf" description="FIFO Configuration"> <fifos> <fifo name="cool_fifo@$${domain}" importance="0"> -  +  </fifo> </fifos> </configuration> An alternate approach is to make the 'member_wait" option an attribute in the member tag, which would be more elegant. However is would require changes to funtions and the fifo_outbound table. I'm sure I could whip that up pretty quickly too. However, I don't want to spend my time doing it without feedback from you all? Please provide some input to the usefulness of this option and and ideas that you have. ./herman freenode: frek310

2 1

[Freeswitch-dev] GSOC09
by Wasim Baig 13 Mar '09

13 Mar '09

Any efforts made to get Freeswitch into the Google Summer of Code 2009? -- wasim h. baig | principal consultant | convergence pk | +92 300 8508070 | peace be upon you ...

3 2

[Freeswitch-dev] Initial build of source on VS 2008
by Chuck Tharp 13 Mar '09

13 Mar '09

Hello, I'm new to FS and trying to build from the source tarball using VS 2008 Express on Windows Vista. Most of the FreeSwitch solution builds successfully, except for some of the sofia support modules. There are many errors, which leads me to believe that I have a missing compiler switch or some other setting. I started by opening the solution file Freeswitch.express.2008.sln. The first few errors are related to the MSG_HEADER_CLASS macro, being called from the sip_basic.c file: ..\..\sofia-sip\libsofia-sip-ua\sip\sip_basic.c(121) : error C2065: 'sip_request_hash' : undeclared identifier ..\..\sofia-sip\libsofia-sip-ua\sip\sip_basic.c(121) : error C2099: initializer is not a constant ..\..\sofia-sip\libsofia-sip-ua\sip\sip_basic.c(121) : error C2065: 'sip_request_d' : undeclared identifier ..\..\sofia-sip\libsofia-sip-ua\sip\sip_basic.c(121) : error C2099: initializer is not a constant ..\..\sofia-sip\libsofia-sip-ua\sip\sip_basic.c(121) : error C2065: 'sip_request_e' : undeclared identifier ..\..\sofia-sip\libsofia-sip-ua\sip\sip_basic.c(121) : error C2099: initializer is not a constant There are many other errors from different source files, but I would guess that they are all related to some setting that I am missing. Thanks, and sorry for the newbie question. Chuck

8 8

[Freeswitch-dev] iptables and traffic control/shaping
by Frank ＠ Impact 12 Mar '09

12 Mar '09

Can anyone point me to any freelancers to do some work on traffic shaping on our linux firewall so we can get our FS voip traffic to flow properly and not compete with our email/web traffic? I have been banging my head and have followed all the details I can find on the FS wiki on this (and about every other web page how to I can find). Still I cannot seem to get it done correctly. Thanks

2 1

[Freeswitch-dev] Possible bugs,
by Szymon Olko 12 Mar '09

12 Mar '09

Hello, Before I open new ticket I want to ask if I should. It happend in svn revision 12293. First occured on kicking member from conference, backtrace is as fallows: #0 switch_core_session_get_channel (session=0x1a) at src/switch_core_session.c:797 #1 0x001bf2e7 in conference_add_event_member_data (member=0x833f1e14, event=0x94ebcc8) at mod_conference.c:417 #2 0x001c81c1 in conf_api_sub_kick (member=0x833f1e14, stream=0xa9b309b8, data=0x0) at mod_conference.c:2936 #3 0x001c253a in conf_api_dispatch (conference=0x9875ad8, stream=0xa9b309b8, argc=3, argv=0xa9b304d8, cmdline=0x8b6eddf7 "agent#385 kick 40441", argn=1) at mod_conference.c:3921 #4 0x001ca4ef in conf_api_main (cmd=0x8b6eddf7 "agent#385 kick 40441", session=0x0, stream=0xa9b309b8) at mod_conference.c:4007 #5 0x0048aa7d in switch_api_execute (cmd=0x8b6eddec "CONFERENCE", arg=0x8b6eddf7 "agent#385 kick 40441", session=0x0, stream=0xa9b309b8) at src/switch_loadable_module.c:1532 #6 0x00a156d2 in api_exec (thread=0x0, obj=0xa9b30a4c) at mod_event_socket.c:1269 #7 0x00a16547 in parse_command (listener=0x93ce980, event=0xa9b31390, reply=0xa9b31123 "", reply_len=512) at mod_event_socket.c:1635 #8 0x00a18283 in listener_run (thread=0xa9bf6b68, obj=0x93ce980) at mod_event_socket.c:2034 #9 0x004e5806 in dummy_worker (opaque=0xa9bf6b68) at threadproc/unix/thread.c:138 #10 0x0037745b in start_thread () from /lib/libpthread.so.0 #11 0x002dde5e in clone () from /lib/libc.so.6 Second was invoked by me when I had zombie conference, conference which has 0 members. I transfered there other member and then called kick all on that conference. backtrace is this: #0 0x0024e3d1 in switch_core_session_perform_kill_channel (session=0x0, file=0x19e7f8 "mod_conference.c", func=0x1a0ac9 "conf_api_sub_kick", line=2929, sig=SWITCH_SIG_BREAK) at src/switch_core_io.c:1048 #1 0x0019813d in conf_api_sub_kick (member=0x67af40bc, stream=0x9f6fb9b8, data=0x0) at mod_conference.c:2929 #2 0x00192654 in conf_api_dispatch (conference=0xa98e73f0, stream=0x9f6fb9b8, argc=3, argv=0x9f6fb4d8, cmdline=0xb297b1ef "agent#341 kick all", argn=1) at mod_conference.c:2731 #3 0x0019a4ef in conf_api_main (cmd=0xb297b1ef "agent#341 kick all", session=0x0, stream=0x9f6fb9b8) at mod_conference.c:4007 #4 0x00258a7d in switch_api_execute (cmd=0xb297b1e4 "conference", arg=0xb297b1ef "agent#341 kick all", session=0x0, stream=0x9f6fb9b8) at src/switch_loadable_module.c:1532 #5 0x00ea46d2 in api_exec (thread=0x0, obj=0x9f6fba4c) at mod_event_socket.c:1269 #6 0x00ea5547 in parse_command (listener=0x9433b80, event=0x9f6fc390, reply=0x9f6fc123 "", reply_len=512) at mod_event_socket.c:1635 #7 0x00ea7283 in listener_run (thread=0xa9161b68, obj=0x9433b80) at mod_event_socket.c:2034 #8 0x002b3806 in dummy_worker (opaque=0xa9161b68) at threadproc/unix/thread.c:138 #9 0x0037745b in start_thread () from /lib/libpthread.so.0 #10 0x005bee5e in clone () from /lib/libc.so.6 I upgrade to trunk SVN of course, but I don't know how to reproduce it so I have to wait. Does anybody recognize similar things?

3 2

[Freeswitch-dev] conference kick all coredump
by Scott Shen 12 Mar '09

12 Mar '09

Dial into a conference, heard announcement 'you are currently the only person in this conference' blah blah Now, in FS console, type conference <conf_name> kick all FS cored immediately. Is that the 'kick all' command shouldn't be used in this way?

4 10

[Freeswitch-dev] C API session and channels questions]
by Chris Danielson 11 Mar '09

11 Mar '09

**this is a repost from the freeswitch-users group. Just wanted to work it out in the dev group. Hello Guys, I have a question regarding how I can do the following within the FS C API. What API can I use to list the current channels and sessions? What I have already in place is a generic FS module capable of serializing any sort of C data and storing it onto another device or caching server. I then have the ability to deserialize and operate on these generic objects already in my module. What I anticipate doing is finding a path to load up the FreeSWITCH (session and channel) objects and deserialize them into a freshly started FreeSWITCH instance. Any ideas on what switch_core_session.c methods, etc. I should use and any pitfalls? My primary focus for my company is conference calling, but I'd like to make this a means to work with everything that I possibly can. As soon as I get my basic testing and instances working I'll offer up the code to bkw, mikej and anthm for review. As-is, I'm not even sure if what I'm asking is even feasible, but I think having a module like this would be a good start to fill my needs and perhaps those of the community. So again, my two questions are as follows: 1) What API can I use to enumerate through the current channels and sessions initialized and running in memory? (I'm looking to clone and serialize them and store them on a separate server). 2) Any helper methods or ways that I can re-construct the channels and sessions into memory on a freshly started instance of FreeSWITCH? Kind Regards, Chris Danielson (aka. danchris)

1 0

[Freeswitch-dev] Why does Freeswitch send Bye to sipp(uac and uas) when doing performance test?
by congxin zhao 10 Mar '09

10 Mar '09

Hi, I use sipp to performance test on freeswitch. The test environment is as following, sipp(uac 10.51.50.5 ) <->(10.51.50.4) freeswitch(10.51.51.4) <-> sipp(uas10.51.51.5 ) But when the new call rate reaches 150 new call per second, after a little while the sipp uas shows the following error. Is there something wrong with my test case? In what environment the freeswitch will send Bye to sipp(uac and uas) ? Thanks -Congxin =============================================================================================================== [czhao@congxin trunk]$ sudo sipp -sn uac -i 10.51.50.5 10.51.50.4 -trace_err -d 100000 -r 200 Resolving remote host '10.51.50.4'... Done. ------------------------------ Scenario Screen -------- [1-9]: Change Screen -- Call-rate(length) Port Total-time Total-calls Remote-host 200.0(100000 ms)/1.000s 5060 51.40 s 10281 10.51.50.4:5060 (UDP) 75 new calls during 0.375 s period 1 ms scheduler resolution 9769 calls (limit 60000) Peak was 9769 calls, after 51 s 0 Running, 9769 Paused, 49 Woken up 0 dead call msg (discarded) 3227 out-of-call msg (discarded) 3 open sockets Messages Retrans Timeout Unexpected-Msg INVITE ----------> 10281 655 0 100 <---------- 10167 214 0 0 180 <---------- 169 0 0 0 183 <---------- 0 0 0 0 200 <---------- E-RTD1 1668 68 0 0 ACK ----------> 1668 0 Pause [ 1:40] 1668 512 BYE ----------> 0 0 0 200 <---------- 0 0 0 0 ------------------------------ Test Terminated -------------------------------- ----------------------------- Statistics Screen ------- [1-9]: Change Screen -- Start Time | 2009-03-09 13:13:49:196 1236575629.196965 Last Reset Time | 2009-03-09 13:14:40:234 1236575680.234329 Current Time | 2009-03-09 13:14:40:609 1236575680.609605 -------------------------+---------------------------+-------------------------- Counter Name | Periodic value | Cumulative value -------------------------+---------------------------+-------------------------- Elapsed Time | 00:00:00:375 | 00:00:51:412 Call Rate | 200.000 cps | 199.973 cps -------------------------+---------------------------+-------------------------- Incoming call created | 0 | 0 OutGoing call created | 75 | 10281 Total Call created | | 10281 Current Call | 9769 | -------------------------+---------------------------+-------------------------- Successful call | 0 | 0 Failed call | 11 | 512 -------------------------+---------------------------+-------------------------- Response Time 1 | 00:00:42:741 | 00:00:18:963 Call Length | 00:00:45:415 | 00:00:31:336 ------------------------------ Test Terminated -------------------------------- 2009-03-09 13:14:40:331 1236575680.331543: Discarding message which can't be mapped to a known SIPp call: INVITE sip:sipp@10.51.50.5:5060 SIP/2.0 Via: SIP/2.0/UDP 10.51.50.4;rport;branch=z9hG4bKa4g2UNXNrty8N Max-Forwards: 70 From: sut <sip:service@10.51.50.4:5060>;tag=Ktmc91a3N33rQ To: sipp <sip:sipp@10.51.50.5:5060>;tag=7557SIPpTag00106 Call-ID: 106-7557(a)10.51.50.5 CSeq: 112169493 INVITE Contact: <sip:mod_sofia@10.51.50.4:5060;transport=udp> User-Agent: FreeSWITCH-mod_sofia/1.0.trunk-11580M Allow: INVITE, ACK, BYE, CANCEL, OPTIONS, PRACK, MESSAGE, SUBSCRIBE, NOTIFY, REFER, UPDATE, REGISTER, INFO Supported: timer, precondition, path, replaces Session-Expires: 120 Min-SE: 120 Content-Type: application/sdp Content-Disposition: session Content-Length: 153 v=0 o=user1 6853881552441804661 3123293097102204886 IN IP4 10.67.7.177 s=- c=IN IP4 10.51.50.4 t=0 0 m=audio 18182 RTP/AVP 0 a=rtpmap:0 PCMU/8000 -------------------------------------------------------------------------------------------------------------------------------------------------- [czhao@congxin trunk]$ sudo sipp -sn uac -i 10.51.50.5 10.51.50.4 -trace_err -d 100000 -r 200 [sudo] password for czhao: Resolving remote host '10.51.50.4'... Done. ------------------------------ Scenario Screen -------- [1-9]: Change Screen -- Call-rate(length) Port Total-time Total-calls Remote-host 200.0(100000 ms)/1.000s 5060 36.35 s 7270 10.51.50.4:5060 (UDP) 66 new calls during 0.326 s period 1 ms scheduler resolution 6915 calls (limit 60000) Peak was 6915 calls, after 36 s 0 Running, 6915 Paused, 0 Woken up 0 dead call msg (discarded) 553 out-of-call msg (discarded) 3 open sockets Messages Retrans Timeout Unexpected-Msg INVITE ----------> 7270 364 0 100 <---------- 7270 123 0 0 180 <---------- 153 0 0 0 183 <---------- 0 0 0 0 200 <---------- E-RTD1 1227 56 0 0 ACK ----------> 1227 0 Pause [ 1:40] 1227 355 BYE ----------> 0 0 0 200 <---------- 0 0 0 0 ------------------------------ Test Terminated -------------------------------- ----------------------------- Statistics Screen ------- [1-9]: Change Screen -- Start Time | 2009-03-09 13:29:10:948 1236576550.948498 Last Reset Time | 2009-03-09 13:29:46:977 1236576586.977287 Current Time | 2009-03-09 13:29:47:307 1236576587.307007 -------------------------+---------------------------+-------------------------- Counter Name | Periodic value | Cumulative value -------------------------+---------------------------+-------------------------- Elapsed Time | 00:00:00:329 | 00:00:36:358 Call Rate | 200.608 cps | 199.956 cps -------------------------+---------------------------+-------------------------- Incoming call created | 0 | 0 OutGoing call created | 66 | 7270 Total Call created | | 7270 Current Call | 6915 | -------------------------+---------------------------+-------------------------- Successful call | 0 | 0 Failed call | 1 | 355 -------------------------+---------------------------+-------------------------- Response Time 1 | 00:00:30:102 | 00:00:13:578 Call Length | 00:00:32:687 | 00:00:23:473 ------------------------------ Test Terminated -------------------------------- 2009-03-09 13:29:47:004 1236576587.004757: Aborting call on an unexpected BYE for call: 673-21803(a)10.51.50.5. sipp: There were more errors, see 'uac_21803_errors.log' file

3 3

[Freeswitch-dev] CCXML/VXML module
by jonathan augenstine 06 Mar '09

06 Mar '09

I have a project that has a requirement for CCXML/VXML. I know this topic has popped up periodically on this list. My first question is, does anyone know of any development going on in this area on Freeswitch? I am considering writing a CCXML/VXML module, but I do not want to reinvent the wheel. If there is no current development going on, then I have some additional questions. I am looking at developing a module. I am familiar with the architecture of Freeswitch, but have not been in the core code a lot. I was considering using the XML-RPC module as a template for a CCMXL/VXML module, and I was considering using expat for XML parsing. Any comments or thoughts? Jonathan

10 19

[Freeswitch-dev] Is there an Opp for FS here in the future?
by EdPimentl 05 Mar '09

05 Mar '09

FYI ------- by I lazaar ..Feb 27, 2009 Back in October of 2007 I authored a column for *Business Communications Review* that looked at Microsoft's entry into the VOIP market and asked whether Microsoft had entered the market after most enterprise organizations had already moved forward with alternate vendors for their VOIP plans, meaning Microsoft was too late to gain significant VOIP market share. At the time of writing, Nemertes Research indicated that 99% of organizations we had interviewed that year had already selected a VOIP vendor and were moving forward with at least a limited pilot, meaning that for Microsoft to succeed in the VOIP arena, it would have to convince potential customers to call a halt to existing plans and switch to Microsoft's Office Communications Server as a VOIP platform. Microsoft made this pitch by promising that it could cut the cost of VOIP in half when compared to its competitors. The customers for the most part didn't listen. The overwhelming majority of participants in our 2008 research interviews said that the only impact of Microsoft's entry into the VOIP space was to pay more attention to the ability to integrate their existing or planned voice systems with the Microsoft Office Communications server and client, enabling features such as "click-to-call" and presence awareness for desktop and Windows Mobile users. Now, about a year and a half later, Microsoft is stepping up its game with the release of OCS Release 2, adding a number of VOIP related features including single-number reachability and dial-in audio conferencing. Perhaps most importantly, OCS R2 adds support for SIP trunking; already Microsoft has announced a number of partnerships with service providers to bring SIP trunks to OCS installations. So with R2 is Microsoft back in the VOIP game? Not yet. Microsoft still faces some of the same issues we heard about in 2007. There's still no good option for remote site survivability nor is there an E-911 capability just yet. Integrating Microsoft into existing voice systems will typically require a mediation server (or multiple servers). Choices for stand-alone desktop phones are limited and most IT architects still don't trust software-based phones running on Windows as their primary means of communication. But maybe we're asking the wrong question? Sure, Microsoft is happy to offer OCS as a phone system replacement to those whose requirements fit Microsoft's capabilities. But perhaps more importantly, OCS is the trojan horse. Today, it offers desktop and mobile device control over communications services, Microsoft's and partners, along with a robust set of collaboration features such as instant messaging, persistent chat, audio, video, and web conferencing, and so on. It integrates presence into the full suite of Microsoft office products including Exchange/Outlook and Office. Web-based clients enable non-Windows users to become part of the Microsoft eco-system. And maybe, just maybe, as OCS's phone feature list grows, and as OCS deployments grow, perhaps with R3 Microsoft will begin to convince a growing number of its customers that they really don't need that old IP-PBX anymore when employees are already using OCS to make the majority of their calls. As we conduct our 2009 round of interviews we're continuing to gather data on the role of Microsoft in VOIP/UC planning. Stay tuned...

1 0