- FreeSWITCH-dev - FreeSWITCH Mailing Lists

Re: [Freeswitch-dev] CCXML/VXML theading
by Anthony Minessale 26 Jun '10

26 Jun '10

Write a native implementation in c or a middleware in event socket? On Jun 25, 2010 10:06 PM, "Jan Berger" <jan.berger(a)video24.no> wrote: Hi, I have a headache with the integration of CCXML/VXML into FreeSWITCH that I hope you might have some input on. The issue is that Call Control and Voice Control in these 2 standards require that Call Control run asynch and process events in parallel with sessions controlling the voice-part of a call. CCXML is event-driven, asynch in nature, while VXML is blocking. Basically that will require a thread for CCXML and a thread for VXML – in reality 2 threads per call. And the threads here are right now Spidermonkey sessions – very expensive. Any clever ideas here would be helpful. Jan _______________________________________________ FreeSWITCH-dev mailing list FreeSWITCH-dev(a)lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-dev UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-dev http://www.freeswitch.org

1 0

CCXML/VXML theading
by Jan Berger 26 Jun '10

26 Jun '10

Hi, I have a headache with the integration of CCXML/VXML into FreeSWITCH that I hope you might have some input on. The issue is that Call Control and Voice Control in these 2 standards require that Call Control run asynch and process events in parallel with sessions controlling the voice-part of a call. CCXML is event-driven, asynch in nature, while VXML is blocking. Basically that will require a thread for CCXML and a thread for VXML - in reality 2 threads per call. And the threads here are right now Spidermonkey sessions - very expensive. Any clever ideas here would be helpful. Jan

1 0

Re: [Freeswitch-dev] uuid_bridge doesn't work (SOLVED)
by Arturo Monroy 24 Jun '10

24 Jun '10

Thaks for your support. regards Arturo Monroy

1 0

FreeSWITCH Conference Call Starting Shortly!
by Michael Collins 23 Jun '10

23 Jun '10

Everyone start calling in! We'll be starting in just a few minutes... http://wiki.freeswitch.org/wiki/FS_weekly_2010_06_23 -MC

1 0

API
by Jan Berger 22 Jun '10

22 Jun '10

Hi folks, I am looking at using ESL to communicate with CCXML and it looks as if I have a challenge as events in FS is a rather short list. The list blow is the CCXML elements & events that must map 1:1 to a telecom platform - I have added the ccxml parameters "as is" - some of these are not needed, but we seem to have a gap between what FS currently export as events and what I will need on ESL. Do we have these events in C? Jan accept ccxml (accept) connectionid hints createconference ccxml (createconference) conferenceid confname reservedtalkers reservedlisteners hints destroyconference ccxml (destroyconference) conferenceid hints disconnect ccxml (disconnect) connectionid reason hints join ccxml (join) id1 id2 duplex hints entertone exittone autoinputgain autooutputgain dtmfclamp toneclamp log ccxml (log) label expr merge ccxml (merge) connectionid1 connectionid2 hints meta ccxml (meta) name http_equiv content metadata ccxml (metadata) redirect ccxml (redirect) connectionid dest reason hints reject ccxml (reject) connectionid reason hints unjoin ccxml (unjoin) id1 id2 hints errorconnectwrongstate ccxml (errorconnectwrongstate) name eventid eventsource eventsourcetype connectionid protocol info connction reason tagname connectiondisconnected ccxml (connectiondisconnected) name eventid eventsource eventsourcetype connectionid protocol info connction reason trigger errorconnection ccxml (errorconnection) name eventid eventsource eventsourcetype connectionid protocol info connction reason connectionaccept ccxml (connectionaccept) name eventid eventsource eventsourcetype connectionid protocol info connction reason connectionfailed ccxml (connectionfailed) name eventid eventsource eventsourcetype connectionid protocol info connction reason connectionredirected ccxml (connectionredirected) name eventid eventsource eventsourcetype connectionid protocol info connction reason connectionredirectfailed ccxml (connectionredirectfailed) name eventid eventsource eventsourcetype connectionid protocol info connction reason connectionrejectfailed ccxml (connectionrejectfailed) name eventid eventsource eventsourcetype connectionid protocol info connction reason connectionprogressing ccxml (connectionprogressing) name eventid eventsource eventsourcetype connectionid protocol info connction connectionconnected ccxml (connectionconnected) name eventid eventsource eventsourcetype connectionid protocol info connction connectionmerged ccxml (connectionmerged) name eventid eventsource eventsourcetype connectionid protocol info connction mergeid connectionmergefailed ccxml (connectionmergefailed) name eventid eventsource eventsourcetype connectionid protocol info connction mergeid reason connectionalerting ccxml (connectionalerting) name eventid eventsource eventsourcetype connectionid protocol info connction connectionsignal ccxml (connectionsignal) name eventid eventsource eventsourcetype connectionid protocol info connction conferencecreated ccxml (conferencecreated) name eventid eventsource eventsourcetype conferenceid conference info conferencedestroyed ccxml (conferencedestroyed) name eventid eventsource eventsourcetype conferenceid info conferencejoined ccxml (conferencejoined) name eventid eventsource eventsourcetype id1 id2 object1 object2 conferenceunjoined ccxml (conferenceunjoined) name eventid eventsource eventsourcetype id1 id2 object1 object2 conferencecreate ccxml (conferencecreate) name eventid eventsource eventsourcetype conferenceid reason errorconferencedestroy ccxml (errorconferencedestroy) name eventid eventsource eventsourcetype conferenceid reason errorconferencejoin ccxml (errorconferencejoin) name eventid eventsource eventsourcetype id1 id2 reason errorconferenceunjoin ccxml (errorconferenceunjoin) name eventid eventsource eventsourcetype id1 id2 reason errorconferenceevent ccxml (errorconferenceevent) name eventid eventsource eventsourcetype conferenceid reason errorsemantic ccxml (errorsemantic) name eventid eventsource eventsourcetype connectionid protocol info connction reason tagname attributelist

2 1

uuid_bridge doesn't work
by Arturo Monroy 22 Jun '10

22 Jun '10

Hello, i have a little problem, could you help me please? I trying to bridge using uuid_bridge API, but doesn't work , and I cant figure out why, I'll explain you: I make a Inboud call and run a lua script code, the lua code has this local UUID = session:getVariable("uuid"); local api = freeswitch.API() local uuid = api:executeString('create_uuid') api:executeString(string.format("originate {ignore_early_media=true,origination_uuid='%s',api_on_answer='uuid_bridge %s %s'}sofia/internal/189(a)192.168.1.203 &park", uuid, UUID, uuid)) The outbpund call is to Xlite phone, and this is ringing but when a answer the call, the first one is hangup and the second (@203) is still alive, What happed?????? The uuid_bridge return +OK but inmediatly the first is over 2010-06-21 12:51:45.725585 [INFO] switch_channel.c:2535 sofia/internal/189(a)192.168.1.203 api on answer: uuid_bridge(7a9ab23d-4ea5-4597-84e2-992ee88c2c79 3456ef2d-ec79-43a7-bf72-a1b2e88fc725) +OK 3456ef2d-ec79-43a7-bf72-a1b2e88fc725 **********Line 765 2010-06-21 12:51:45.725585 [DEBUG] switch_core_state_machine.c:344 (sofia/internal/189(a)192.168.1.203) State RESET going to sleep 2010-06-21 12:51:45.725585 [DEBUG] switch_ivr_originate.c:3341 Originate Resulted in Error Cause: 487 [ORIGINATOR_CANCEL] ********* Line 780 Please help me, thank you -- Saludos Arturo Monroy

4 5

call legs
by vectorno vectorno 21 Jun '10

21 Jun '10

How can the call leg be identified on a channel hangup

1 0

Hook-Flash passthrough/translation FXS->FXO
by Jeroen C. van Gelderen 21 Jun '10

21 Jun '10

Hi Guys, [In short: can someone knowledgeable confirm the small patch below for hook-flash support uses correct locking? Is there a better approach?] I am taking 16 CENTREX (analog) lines from our telco and passing each through a Freeswitch instance to 16 analog handsets in order to enable call recording. I'm using a Xorcom device with 16 FXS and 16 FXO ports and I'm basically bridging these ports together one-to-one. Since this setup exists side-by-side with 100 "unadorned" CENTREX lines it is important that the clerks don't have to learn a second way of transferring calls for some percentage of the stations. Calls on CENTREX lines are transferred by pressing <hook flash> # <digit> <digit> (i.e. <flash>#42) DTMF digits can be made to pass through in-band just fine (modulo a feedback issue that I'm still working on but have patched around for now) but the hook flash support is only half-way present. Freeswitch already supports the artificial DTMF digit "F" will allow you to queue it on a channel to effect a hook flash. It does not yet allow detection of the hook-flash on the FXS to result in digit "F" being queued. I'm running with the following (hack alert) patch to make this work. I'm considering tying this to a knob called flash2dtmf or similar. ----8<----8<----8<----8<----8<----8<----8<---- @@ -1746,6 +1750,31 @@ static ZIO_SIGNAL_CB_FUNCTION(on_fxs_signal) zap_set_flag_locked(sigmsg->channel, ZAP_CHANNEL_HOLD); zap_set_state_locked(sigmsg->channel, ZAP_CHANNEL_STATE_DIALTONE); } + } else { + switch_core_session_t *session; + switch_dtmf_t _dtmf = { 'F', SWITCH_DEFAULT_DTMF_DURATION }; + + zap_log(ZAP_LOG_DEBUG, "HERE BE CUSTOM FLASH HANDLING...\n"); + + if ((session = zap_channel_get_session(sigmsg->channel, 0))) { + + zap_log(ZAP_LOG_DEBUG, "Got a session...\n"); + + channel = switch_core_session_get_channel(session); + + zap_log(ZAP_LOG_DEBUG, "Got a channel...\n"); + + switch_channel_queue_dtmf(channel, &_dtmf); + + zap_log(ZAP_LOG_DEBUG, "Queued DTMF [F]...\n"); + + switch_core_session_rwunlock(session); + + zap_log(ZAP_LOG_DEBUG, "Unlocked session...\n"); + } + + zap_log(ZAP_LOG_DEBUG, "END OF CUSTOM FLASH HANDLING.\n"); + } } ----8<----8<----8<----8<----8<----8<----8<---- A log excerpt of a clerk taking an inbound call (14sec) and then using hook-flash to transfer the call to extension #42: ----8<----8<----8<----8<----8<----8<----8<---- 2010-06-15 18:12:33.435334 [NOTICE] mod_openzap.c:1612 Channel [OpenZAP/1:1/] has been answered 2010-06-15 18:12:33.451333 [DEBUG] switch_core_codec.c:146 OpenZAP/3:1/1001 Restore previous codec PCMU:0. 2010-06-15 18:12:33.451333 [DEBUG] switch_ivr_originate.c:3230 Originate Resulted in Success: [OpenZAP/1:1/] 2010-06-15 18:12:33.451333 [DEBUG] switch_ivr_originate.c:3230 Originate Resulted in Success: [OpenZAP/1:1/] 2010-06-15 18:12:33.451333 [DEBUG] switch_core_session.c:643 Send signal OpenZAP/1:1/ [BREAK] 2010-06-15 18:12:33.451333 [DEBUG] switch_core_session.c:643 Send signal OpenZAP/3:1/1001 [BREAK] 2010-06-15 18:12:33.451333 [DEBUG] switch_ivr_bridge.c:1182 (OpenZAP/1:1/) State Change CS_CONSUME_MEDIA -> CS_EXCHANGE_ MEDIA 2010-06-15 18:12:33.451333 [DEBUG] switch_core_session.c:1023 Send signal OpenZAP/1:1/ [BREAK] 2010-06-15 18:12:33.451333 [DEBUG] switch_core_state_machine.c:314 (OpenZAP/1:1/) Running State Change CS_EXCHANGE_MEDIA 2010-06-15 18:12:33.451333 [DEBUG] switch_core_state_machine.c:351 (OpenZAP/1:1/) State EXCHANGE_MEDIA 2010-06-15 18:12:33.451333 [DEBUG] mod_openzap.c:595 CHANNEL EXCHANGE_MEDIA 2010-06-15 18:12:33.455333 [DEBUG] switch_core_session.c:704 Send signal OpenZAP/1:1/ [BREAK] 2010-06-15 18:12:33.455333 [DEBUG] switch_core_session.c:704 Send signal OpenZAP/3:1/1001 [BREAK] 2010-06-15 18:12:47.543864 [DEBUG] ozmod_analog.c:801 EVENT [FLASH][1:1] STATE [UP] 2010-06-15 18:12:47.543864 [DEBUG] mod_openzap.c:1605 got FXS sig [FLASH] 2010-06-15 18:12:47.543864 [DEBUG] mod_openzap.c:1757 HERE BE CUSTOM FLASH HANDLING... 2010-06-15 18:12:47.543864 [DEBUG] mod_openzap.c:1761 Got a session... 2010-06-15 18:12:47.543864 [DEBUG] mod_openzap.c:1765 Got a channel... 2010-06-15 18:12:47.543864 [DEBUG] mod_openzap.c:1769 Queued DTMF [F]... 2010-06-15 18:12:47.543864 [DEBUG] mod_openzap.c:1773 Unlocked session... 2010-06-15 18:12:47.543864 [DEBUG] mod_openzap.c:1776 END OF CUSTOM FLASH HANDLING. 2010-06-15 18:12:47.570863 [DEBUG] zap_io.c:2062 3:1 GENERATE DTMF [F] 2010-06-15 18:12:49.554796 [DEBUG] mod_openzap.c:724 queue DTMF [#] 2010-06-15 18:12:49.570795 [DEBUG] zap_io.c:2062 3:1 GENERATE DTMF [#] 2010-06-15 18:12:49.774789 [DEBUG] mod_openzap.c:784 Dropping frame! (write not ready) 2010-06-15 18:12:50.194774 [DEBUG] mod_openzap.c:724 queue DTMF [4] 2010-06-15 18:12:50.230773 [DEBUG] zap_io.c:2062 3:1 GENERATE DTMF [4] 2010-06-15 18:12:50.314771 [DEBUG] mod_openzap.c:722 Ignoring DTMF [4] on FXO port 3:1 2010-06-15 18:12:50.414767 [DEBUG] mod_openzap.c:784 Dropping frame! (write not ready) 2010-06-15 18:12:50.794755 [DEBUG] mod_openzap.c:724 queue DTMF [2] 2010-06-15 18:12:50.830753 [DEBUG] zap_io.c:2062 3:1 GENERATE DTMF [2] 2010-06-15 18:12:50.915750 [DEBUG] mod_openzap.c:722 Ignoring DTMF [2] on FXO port 3:1 2010-06-15 18:12:51.013748 [DEBUG] mod_openzap.c:784 Dropping frame! (write not ready) 2010-06-15 18:13:29.244471 [DEBUG] ozmod_analog.c:801 EVENT [ONHOOK][1:1] STATE [UP] 2010-06-15 18:13:29.244471 [DEBUG] ozmod_analog.c:837 Changing state on 1:1 from UP to DOWN 2010-06-15 18:13:29.262471 [DEBUG] ozmod_analog.c:463 Executing state handler on 1:1 for DOWN 2010-06-15 18:13:29.262471 [DEBUG] mod_openzap.c:1605 got FXS sig [STOP] 2010-06-15 18:13:29.262471 [DEBUG] switch_channel.c:2257 (OpenZAP/1:1/) Callstate Change ACTIVE -> HANGUP 2010-06-15 18:13:29.262471 [NOTICE] mod_openzap.c:1696 Hangup OpenZAP/1:1/ [CS_EXCHANGE_MEDIA] [NORMAL_CLEARING] 2010-06-15 18:13:29.262471 [DEBUG] switch_channel.c:2273 Send signal OpenZAP/1:1/ [KILL] 2010-06-15 18:13:29.262471 [DEBUG] switch_core_session.c:1023 Send signal OpenZAP/1:1/ [BREAK] ----8<----8<----8<----8<----8<----8<----8<---- Cheers, -Slim -- Jeroen C. "Slim" van Gelderen Olympic Sports Data Services Email: jeroeng(a)thegreek.com Phone: +1 876 953 6182 x128

3 4

mod_limit / db / hash changes - If you update action is required
by Rupa Schomaker 21 Jun '10

21 Jun '10

If you've been paying attention to the conf call or irc you'll know this change was coming. I've committed a significant change to how limits are managed in FreeSWITCH. This change is NOT backwards compatible. It requires one to: 0) git pull (of course) 1) rerun ./configure to generate new Makefiles for new modules 2) edit modules.conf, remove mod_limit and add mod_db and mod_hash 3) modify conf/autoload_configs/modules.conf.xml, remove mod_limit and add mod_db and mod_hash 4) create/edit conf/autoload_configs/db.conf.xml if using ODBC In your dialplan, anywhere you are using limit, you need to add db to the data line. Anywhere you are using limit_hash you need to remove _hash from application and add hash to the front of the data line. Same pattern applies to any API usage of limit you may be using. Read the wiki at: http://wiki.freeswitch.org/wiki/Limit for documentation on limit. Ok, why the change? limit has been moved into core with support for pluggable backend limit providers. This allows one to develop limit backends relatively easily and provides a more structured interface to these different limit backends. Expect to see backends provided by more than just db and hash. It also always bothered me that db and hash were part of the mod_limit module. They are not standalone and provide functionality specific to themselves. I've updated the wiki with documentation. I'll continue to update the wiki as necessary, feel free to do so yourself if there is something unclear or missing. -- -Rupa

2 2

mod_lcr update - read if you use custom_sql
by Rupa Schomaker 20 Jun '10

20 Jun '10

mod_lcr has been updated. commit log: mod_lcr update. MODAPP-340, MODAPP-355 arbitrary b-leg vars, limit support, lcr/ endpoint custom_sql now is field based rather than position based NOTE: custom_sql is incompat with prior version arbitrary b-leg vars enable integration with mod_nibblebill. Wiki has been updated: http://wiki.freeswitch.org/wiki/Mod_lcr This change is backwards compatible EXCEPT if you use custom_sql. If you use custom_sql you must update your sql to use field aliases (SQL : AS). The module no longer uses position in the custom sql for fields. So, if you use custom_sql, reread: http://wiki.freeswitch.org/wiki/Mod_lcr#Custom_SQL Ok, I'm done breaking things for the day. :) -- -Rupa

1 0