September 2012 - FreeSWITCH-users - FreeSWITCH Mailing Lists

Re: [Freeswitch-users] sip-router or opensips for good
by Yufei Tao 12 Sep '12

12 Sep '12

If you use TCP/TLS a lot, Kamailio will save you a lot of grief as it's got a much better TCP stack. I also find the cfg file is much easier to work with as already mentioned. Yufei On 29/08/12 11:02, freeswitch-users-request(a)lists.freeswitch.org wrote: > From: freeswitch-users-bounces(a)lists.freeswitch.org > > [mailto:freeswitch-users-bounces@lists.freeswitch.org] On Behalf Of > > Adam Kelloway > > Sent: Thursday, August 23, 2012 1:55 PM > > To: FreeSWITCH Users Help > > Subject: Re: [Freeswitch-users] sip-router or opensips for good > > interoperability with freeswitch? > > > > > > > > > > The are both pretty well functionally the same. You might find that > > Kamailo's .cfg file is easier to work with, as it allows you to name > > the routes (among other things), and it has a decent default cfg to > > use as a starting point. I have known people to have success with both > > opensips and kamailio. -- Yufei Tao Red Embedded This E-mail and any attachments hereto are strictly confidential and intended solely for the addressee. If you are not the intended addressee please notify the sender by return and delete the message. You must not disclose, forward or copy this E-mail or attachments to any third party without the prior consent of the sender. Red Embedded Design, Company Number 06688253 Registered in England: The Waterfront, Salts Mill Rd, Saltaire, BD17 7EZ

4 3

FreeSwitch crashes when called party hangs up on Prompt Play
by Adelia C. 12 Sep '12

12 Sep '12

Problem: FreeSwitch crashes when called party hangs up on prompt played (prompt played on called side). Interestingly enough, this does not happen in DEV (win32 Windows 2003 VMs) but happens on win32 Windows 2003 HW (real hardware). Offending lib is libapr.dll. I ran libapr.dl through Dependency Walker on both VM and HW and all is equal. FreeSwitch version : GIT version from 2 weeks ago, Release version only (no Debug on DEV) Calling Function at crash time (full trace at http://pastebin.freeswitch.org/19876) : 2012-09-11 15:38:09.534355 [NOTICE] switch_cpp.cpp:1227 8b24ba68-79ea-4ab3-9d22-ae3da858f1a1 Proxying Call routes are: {origination_caller_id_number=4152484093,origination_uuid=663ec186-36ed-4f40-8e68-4dad5d8550eb,ignore_early_media=true,group_confirm_key=exec,group_confirm_file=playback //ashstgfs.stage.keen.com/files/Callcloud/NewWhisper_d5d694a0_a402_491e_a5c2_ac15c8c5b92f.wav}sofia/gateway/26/4152484093 | [origination_uuid=a9dfe5b9-bdd9-40d9-8a2b-897c5c164825]sofia/gateway/25/4152484093 | [origination_uuid=c7cb97af-1dbb-4c84-81be-c8e794000a27]sofia/gateway/7/4152484093 | [origination_uuid=1612b58e-86e7-45e2-bea3-b81360139fea]sofia/gateway/10/4152484093 6815a8a9-fb0b-41dd-b2d0-cabfe7953ab4 EXECUTE sofia/external/4152484093(a)10.3.220.50 bridge({origination_caller_id_number=4152484093,origination_uuid=663ec186-36ed-4f40-8e68-4dad5d8550eb,ignore_early_media=true,group_confirm_key=exec,group_confirm_file=playback //ashstgfs.stage.keen.com/files/Callcloud/NewWhisper_d5d694a0_a402_491e_a5c2_ac15c8c5b92f.wav}sofia/gateway/26/4152484093 | [origination_uuid=a9dfe5b9-bdd9-40d9-8a2b-897c5c164825]sofia/gateway/25/4152484093 | [origination_uuid=c7cb97af-1dbb-4c84-81be-c8e794000a27]sofia/gateway/7/4152484093 | [origination_uuid=1612b58e-86e7-45e2-bea3-b81360139fea]sofia/gateway/10/4152484093) 6815a8a9-fb0b-41dd-b2d0-cabfe7953ab4 2012-09-11 15:38:09.549980 [DEBUG] switch_channel.c:1067 sofia/external/4152484093(a)10.3.220.50 EXPORTING[export_vars] [RFC2822_DATE]=[Tue, 11 Sep 2012 15:38:09 Pacific Daylight Time] to event Exception: Event Type: Error Event Source: .NET Runtime Event Category: None Event ID: 1026 Date: 9/11/2012 Time: 3:38:31 PM User: N/A Computer: ASHSTGSIP02 Description: Application: FreeSwitchConsole.exe Framework Version: v4.0.30319 Description: The process was terminated due to an unhandled exception. Exception Info: System.AccessViolationException Stack: at CoreSession.execute(CoreSession* , SByte* , SByte* ) at CSharp_CoreSession_Execute(Void* jarg1, SByte* jarg2, SByte* jarg3) at FreeSWITCH.Native.freeswitchPINVOKE.CoreSession_Execute(HandleRef jarg1, String jarg2, String jarg3) at FreeSWITCH.Native.CoreSession.Execute(String app, String data) at CallCloudAppServer.Service.Service.ProxyInboundCall(String rtn, String callerIdNumber, Boolean callRecordingRoute) at CallCloudAppServer.Service.CallTracking.CallTrackingService.DoVanillaCallTracking() at CallCloudAppServer.Service.CallTracking.CallTrackingService.ProcessIncomingCall() at CallCloudAppServer.Core.Dispatcher.Dispatch() at CallCloudAppServer.Core.MessageHandler.MessageHandler.ProcessIncomingCall() at CallCloudAppServer.Main.Run(AppContext context) at FreeSWITCH.AppPluginExecutor.Execute(String args, IntPtr sessionHandle) at FreeSWITCH.AppPluginExecutor.Execute(System.String, IntPtr) at FreeSWITCH.Loader.Run(System.String, IntPtr) Cause : passing in wrong parameters to triggers Access Violation - 0xc0000005 - in libapr!apr_socket_sendto. LOGS and CRASH : LOGS at http://pastebin.freeswitch.org/19876 CRASH.DMP at http://pastebin.freeswitch.org/19877 Any idea? Thanks for your help! A.C.

2 3

Work around for Chrome and Flex Phone
by Jack 12 Sep '12

12 Sep '12

Chrome can be tweaked to work with the Flex phone on conference.freeswitch.org by going to the plugins ( chrome://plugins ) and disabling pepflashplayer.dll. Chrome Canary works without tweaking.

1 0

Note: ./configure with --enable-64 on 32-bit platform gives "error: no usable zlib; please install zlib devel package or equivalent"
by Zenaan Harkness 12 Sep '12

12 Sep '12

As per subject, ./configure when run with --enable-64 on 32-bit Debian testing gives me "error: no usable zlib; please install zlib devel package or equivalent". Perhaps it would be good to have a ./configure warning or something - I guess it's conceivable that someone might want to cross compile _to_ amd64, _from_ a 32-bit platform (is this even possible?) but I'd assume would generally be an error. This email was going to be a plea for help/ assumed bug, before I (thankfully, and saving me from a brown paper bag) realised I was copying my 64-bit command line into my 32-bit computer's remote session ... The erroneous (from my perspective) output had me try a few pathways over the last day+ (yeah yeah, brown paper bag), before I just then realised (don't rub it in) that --enable-64 is perhaps not quite ideal on 32 bit hardware. (Aside: for the record, I've tried testing freeswitch and client on same (dual core) laptop: linphone won't log in (sip) and only says "port 5060 may already be in use" or somesuch" - perhaps it's trying to run as a server?; jitsi works but really rough latency, huge mouse lags and jumps etc (yet to test with ..-rt real time debian kernel); qutecom appears to work very well so far (5s audio echo and space invaders soundtrack tests). ) zen

1 0

channels not in sync on stereo recording
by Dave Horton 12 Sep '12

12 Sep '12

I'm experiencing a strange problem on a recording of an audio stream where the two channels (I am recording in stereo) are not in sync. By that, I mean when I play back the audio one of the tracks is somehow time-shifted, so that it doesn't match up to the proper place in the timeline with the other side. For instance, if I call out to a bank IVR system that plays a prompt, and I barge into the prompt by pressing a DTMF key, on the recording I will hear the prompt start and then stop at the same place it stopped when I pressed the DTMF, but the DTMF key itself does not show up on the recording for 5 seconds more. Has anyone ever heard of anything like this? Now, a few more details. I am creating a bridge conference and then I am recording the audio stream to and from the called party and the conference. I do not have this problem if I record the audio stream to and from the calling party and the conference. Here is the dial plan I am using: <extension name="6174973895" continue="false"> <condition field="source" expression="mod_sofia"/> <condition field="destination_number" expression="6174973895"> <action application="ring_ready"/> <action application="answer"/> <action application="start_dtmf"/> <action application="set" data="call_id=${strftime(%Y%m%d_%H%M%S)}_${sip_from_tag}"/> <action application="export" data="outfile=$${base_dir}/tmp/${call_id}.wav"/> <action application="lua" data="demo_call2.lua"/> <action application="export" data="RECORD_STEREO=true"/> <action application="export" data="nolocal:execute_on_media=record_session ${outfile}"/> <action application="export" data="nolocal:api_on_answer=conference ${call_id} play misc/this-call-may-be-recorded.wav async"/> <action application="conference" data="bridge: ${call_id}@simple:{origination_caller_id_number=${caller_id_number}}sofia/normal_customer/${outdial}@${egress_gateway}"/> </condition> </extension> Any thoughts on what might be going on?

1 0

Callers are put into queue although there are free phones
by Fahrwerk Kurierkollektiv (Technik) 12 Sep '12

12 Sep '12

Hey, we're running a freeswitch instance in our small company on a Linux HA Cluster under Debian. Since we are a bicycle messenger company, a correctly working telephone system is absolutely crucial. We use a total of four SNOM Phones (300 series), two of which are "incoming" for calling customers and the remaining two are used to communicate with our couriers, e.g. in case that the radio doesn't work. Of lately, some customers reported to us, that they have been put into our queue for more than 10 minutes until they hang up. A normal call from or to a customer is not longer than 4 minutes and in most cases we are sure that there was absolutely no reason to put the person into the queue because both phones were free at the time of calling. A collegue of mine proposed that it might have to do something with the following advice taken from the wiki: "If calls are not properly ended by freeswitch after the phone has hung up, check if the BYE messages sent by the phone are authenticated. (containing a proxy-authentication line) If not, you can work around this by adding <param name="auth-all-packets" value="false"/> to the sip-profile." http://wiki.freeswitch.org/wiki/Interop_List#SNOM_190.2C_snom300_series.2C_… (The callers number is 493027574754) I have posted a freeswitch log to pastebin of one confirmed call where the customer was put into queue although none of the phones was occupied: http://pastebin.com/NdVZ9DdS Also here's how we set up our queue, which is fairly simple I guess: http://pastebin.com/sGNBmk2d I can, of course, provide more configuration snippets and so on, if requested and hope that you can help us fix the problem :) Greets, Markus

1 1

link error when compiling mod_dingaling
by Carl Johan Jensen 12 Sep '12

12 Sep '12

I have cloed the V1.2.stable and tries to compile it using Visual C++ 2008 I get the below linking error (and several more) What am I missing to include / do ? 3> Creating library Win32\Debug/mod_dingaling.2008.lib and object Win32\Debug/mod_dingaling.2008.exp 3>iksemel.lib(stream.obj) : error LNK2019: unresolved external symbol _SSL_free referenced in function _handshake 3>iksemel.lib(stream.obj) : error LNK2019: unresolved external symbol _SSL_connect referenced in function _handshake Best Regards Carl Johan Jensen

1 0

Missing RPM repo archive
by ＠gbraad 12 Sep '12

12 Sep '12

Hi, The file http://files.freeswitch.org/yum/freeswitch-release-1-0.noarch.rpm seems to be missing, so for now you would need to create your own repo file in /etc/yum.repos.d/ as freeswitch.repo containing the following: [freeswitch] name=FreeSwitch RPMs for Enterprise Linux $releasever - $basearch baseurl=http://files.freeswitch.org/yum/$releasever/$basearch/ enabled=1 gpgcheck=0 Take note of the gpgcheck=0, since the key is not available for verification. Please, provide the needed repo archive... regards, Gerard -- Gerard Braad — 吉拉德 F/OSS & IT Consultant in Beijing http://gbraad.nl gpg: 0x592CFE75

2 2

t38 issue
by royj 11 Sep '12

11 Sep '12

Hi all There is some strange issue with t38-passthru (FreeSWITCH Version 1.2.0-rc2). Task - t38-passthru spa112(ip 1.1.1.1) <--(profile local)--> FreeSWITCH(ip 2.2.2.2)<--(profile extrernal)--> Mediant 2000(ip 3.3.3.3); fax from spa112 to Mediant 2000. In both profiles: <param name="t38-passthru" value="true"/> <param name="inbound-proxy-media" value="true"/> <param name="inbound-late-negotiation" value="true"/> Here is the sip-trace of the call with fax - http://pastebin.freeswitch.org/19866 It is seen that t38 re-INVITE from spa112 (profile local) passed on and back with "200 OK", but t38 from Mediant 2000 (profile external) is not (in response to re-INVITE t38 (SDP, m=image) - "200 OK" with SDP m=audio). After agreement the SDP, the data (v21-preamble) is not passed by freeswitch between spa112 and Mediant 2000, and one of the party ends conversation with message "no-signal" and then "by" . Wiki says ( http://wiki.freeswitch.org/wiki/Proxy_Media#Why_you_almost_certainly_don.27… ) that module mod_spandsp is required (if I understand correctly), but "load mod_spandsp" does not change anything. I would appreciate your help -- Regards royj

3 2

best FS interface for development
by Brett Clark - Grasshopper 11 Sep '12

11 Sep '12

There seems to be at least 4 general ways for customizing FS: - Use the event socket and build an app to interface with it - Use mod lua, or similar, to develop in a particular language - Develop your own 'module' to interface directly with the core - Use the XML interface to implement IVR, PBX, and voicemail type functionality and all the rest. How prevalent is each approach in the community? It seems like everyone is doing a little of both, but maybe someone has a better handle of what the actual ratios are? Which is best suited for an arbitrarily large and complex application? I realize this isn't a simple question-what I want is to understand the most evolved and featureful way to interface with FS? I don't want to adopt an approach which isn't being actively maintained or is missing features. As new stuff is added to FS, which approach will allow me to adopt those new features most easily? Thanks! Brett

8 10