August 2011 - FreeSWITCH-users - FreeSWITCH Mailing Lists

freetdm error message
by ovvenkat 03 Aug '11

03 Aug '11

Hi to all, In my "fs_cli" I am getting bellow message very frequently. Can you any one help me where is the error? 2011-08-02 17:11:33.028452 [INFO] ftmod_sangoma_isdn_stack_rcv.c:995 sng_isdn->s1: Invalid Q.921/Q.931 frame - ignoring len:1 2011-08-02 17:11:33.048455 [INFO] ftmod_sangoma_isdn_stack_rcv.c:995 sng_isdn->s1: Invalid Q.921/Q.931 frame - ignoring len:1 2011-08-02 17:11:33.048455 [INFO] ftmod_sangoma_isdn_stack_rcv.c:995 sng_isdn->s1: Invalid Q.921/Q.931 frame - ignoring len:1 2011-08-02 17:11:33.068455 [INFO] ftmod_sangoma_isdn_stack_rcv.c:995 sng_isdn->s1: Invalid Q.921/Q.931 frame - ignoring len:1 2011-08-02 17:11:33.068455 [INFO] ftmod_sangoma_isdn_stack_rcv.c:995 sng_isdn->s1: Invalid Q.921/Q.931 frame - ignoring len:1 2011-08-02 17:11:33.068455 [INFO] ftmod_sangoma_isdn_stack_rcv.c:995 sng_isdn->s1: Invalid Q.921/Q.931 frame - ignoring len:1 -- Regards Venkatesan OV.

3 3

conference caller id number&name
by ran zhang 03 Aug '11

03 Aug '11

I want when the bridging conference is inviting someone to start the conference, the caller is shown as the real caller id number&name rather than whats stored in 'caller-id-number' and 'caller-id-name' in conference.conf.xml.

2 1

TTS/unimrcp engines
by Jason Moran 02 Aug '11

02 Aug '11

I'm trying to wrap up which TTS engine we will use over unimrcp. We'll use between 10 and 40 simultaneous TTS lines. So far I have tested the new IVONA mrcpv1 plugin but I am waiting on their dev's to work out a few issues I uncovered. Which TTS engines have been successfully used by others? I've heard good things about Loquendo, but I have had problems getting in contact with them. Acapela sounds nice on their website - does anybody know how their unimrcp plugin sounds? Cepstral is cheap, but it seems as though they do not have a current release for Open Suse 11.6 (32bit). Any others? Recommendations? -Jason

2 3

Infolink, ServerPronto, ColoPronto, Etc <YES ITS IS OFF TOPIC>
by Ken Rice 02 Aug '11

02 Aug '11

Just a word of warning... There is a company in Florida that operates under many company names including but not limited to InfoLink, ServerPronto, ColoPronto, and others... This Company makes it IMPOSSIBLE to close an account... I had colocation (rented a dedicated server) from them for several months, during the entire time, the bandwidth was never right, altho they promised several TB of transfer, bandwidth was such that promised transfer amounts were never achievable... I secured separate Colo, issued an order to cancel the account by opening a ticket through their secured web portal... These idiots continue to attempt to charge my credit card to the point where I had to cancel my card and order a new one... And til this day they continue to try and collect and refuse to refund any money they collected after the cancel order... Avoid these guys like the plague unless you want to jump through hoops K

1 0

mod_rtmp question
by Madovsky 02 Aug '11

02 Aug '11

I use freeswitch in ODBC mode. in sofia endpoint if I type "sofia status profile internal" I get all registrations from all nodes but in rtmp endpoint "rtmp status profile internal reg" I get registrations of the local node. is it a normal behaviour ? if yes which rtmp command can give the same result of sofia with ODBC ? Thanks Franck

2 2

How to use ESL event
by David Ma 02 Aug '11

02 Aug '11

Hello All, I have some problem with receiving and handling ESL events. My code is like: void handle_eslswitch_event_plain(esl_handle_t *handle, esl_event_t *event) { ... ... TRACE_DEBUG("ESL event [%d] %s: UUID: %s", event->event_id, fs_get_event_header(event, "Event-Name"), uniqueid); print_event(event); ... ... } void handle_event(esl_handle_t *handle, esl_event_t *last_event) { ... ... if (!strcasecmp(type, "text/event-plain")) { handle_eslswitch_event_plain(handle, handle->last_ievent); } ... ... } int main() { ... ... //receive from event socket if (handle.last_event) handle_event(&handle, handle.last_event); ... ... Usually I receive correct events. But I occasionally receive incorrect event. See below the 1st is the normal event and the 2nd is the wrong event: <2011-07-29 14:25:29> [DEBUG] ESL event [8] CHANNEL_HANGUP_COMPLETE: UUID: f320e4a6-db23-46d0-8d89-9957eccbd4c9 <2011-07-29 14:25:29> [DEBUG] RECV EVENT Event-Name: CHANNEL_HANGUP_COMPLETE Core-UUID: 26b77cba-8fdd-486d-90ec-6844bca58c72 FreeSWITCH-Hostname: fs01 FreeSWITCH-IPv4: 10.1.1.46 FreeSWITCH-IPv6: ::1 Event-Date-Local: 2011-07-29 14:25:29 However, usually after failure in executing "hangup" and getting "-ERR ..." in the event->last_sr_reply, I have wrong event, but not always (an -ERR returned from executing "hangup" doesn't always result in a wrong event). Here it is: <2011-07-29 14:25:29> [NOTICE] -ERR invalid session id [f320e4a6-db23-46d0-8d89-9957eccbd4c9] <2011-07-28 15:55:23> [DEBUG] ESL event [0] : UUID: <2011-07-28 15:55:23> [DEBUG] RECV EVENT Content-Length: 6485 Content-Type: text/event-plain Event-Name: CHANNEL_HANGUP Core-UUID: 26b77cba-8fdd-486d-90ec-6844bca58c72 FreeSWITCH-Hostname: fs01 FreeSWITCH-IPv4: 10.1.1.46 FreeSWITCH-IPv6: %3A%3A1 Event-Date-Local: 2011-07-28%2015%3A55%3A23 In the above case, both Event ID and UUID are invalid. I think i need more information regarding how to use the data elements in ESL event, especially how to use the following: char last_reply[1024]; /*! Las command reply when called with esl_send_recv */ char last_sr_reply[1024]; /*! Last event received. Only populated when **save_event is NULL */ esl_event_t *last_event; /*! Last event received when called by esl_send_recv */ esl_event_t *last_sr_event; /*! This will hold already processed events queued by esl_recv_event */ esl_event_t *race_event; /*! Events that have content-type == text/plain and a body */ esl_event_t *last_ievent; My questions, what event should I check if I am interested in only channel/call related events, and how? Do I check handle.last_event, or handle.last_ievent, or what? Please kindly advice. Thanks, D.Ma

1 0

ESL: not bein able to determine when SIP gateway is down when originating a call
by Anton VG 02 Aug '11

02 Aug '11

I'm trying to catch an error, in case I would dial wrong (non existent) gateway (intentionally!) I'm running ESL outbound listener, subscribing to all events, if I do bgapi 'originate' to a live gateway - there are normal events flow, and I can track what is happening. But if I issue originate to a gateway, which is not configured or simply down - there are no any events fired. I only have an error on FS console 2011-05-14 20:58:13.072927 [ERR] mod_sofia.c:4044 Invalid Gateway following by 2011-05-14 20:58:13.072927 [ERR] switch_ivr_originate.c:2447 Cannot create outgoing channel of type [sofia] cause: [INVALID_NUMBER_FORMAT] But HOW to catch the given in ESL? sofia.c seems just does not have event code for that cases if (profile_name && !profile_found) { switch_log_printf(SWITCH_CHANNEL_LOG, SWITCH_LOG_WARNING, "No Such Profile '%s'\n", profile_name); >status = SWITCH_STATUS_FALSE; } logically there should be a proper way to determine that gateway is bad in my ESL dialplan, by catching the proper event/reply/whatever, For the moment i did trick: esl.api('sofia status gateway GatewayWhichIsDown') When in production, and there is more than a single route, there will be plenty of cases, when you dial a bad gateway, so there should be a way for ESL dialplan to determine that a gateway is not callable for a moment, the reason WHY and to retry with another one. The trick above is bad, since: 1. blocking api query, before evey single gateway call attempt. 2. Gateway maybe known in UP state, but the state is stale, in dial in fact will go to DOWN gateway. So, ESL dialplan will screw in that case Any clue?

4 4

Various question - CallerID / Call transfer via REFER and CDR
by Weigel, Stefan 02 Aug '11

02 Aug '11

Hi all, I'm currently stuck with some issues I can't figure out. I'm using FreeSWITCH Version 1.0.head (git-db5f504 2011-07-17 17-00-38 -0400), phone devices are Polycom Soundpoint IP 560. 1. When receiving a call I change some of the incoming Caller-ID names. It's working like a charm. But if I do a call forward (it's a 'Consultative transfer'-Party A dials party C's number and talks privately with party C after the call is answered, and then completes the transfer or hangs up) I only see the party A's number/name on the display C. Is there a possibility to get the original caller-ID/-name displayed on the phone of Party C ? 2. When doing a 'Consultative transfer', I only get one CDR entry using mod_cdr_csv. For example calling from my cellphone 0179 12345678 to extension 615, then doing a call transfer from 615 to 803 I get an entry in CDR with calling_id = 615, destination_number = 803. Is it possible to trigger the creation of a CDR entry ? I tried using process_cdr but only one entry get's logged. In cdr_csv.conf I'm using logging of a-leg, for billing / call statistic. Thanks in advance and best regards Stefan

1 0

Re: [Freeswitch-users] Sip Headers advice (Not parsing properly)
by Dissident 02 Aug '11

02 Aug '11

Problem solved!!! It seems that this doc is misleading -> http://wiki.freeswitch.org/wiki/Channel_Variables#Channel_Variable_Manipula… not all sip headers are gonna be parsed only the ones which start with X- or P- the rest are not parsed because they are not supposed to exist, however in my case they do and the only solution I've found to parse my CISCO headers is modifying the mod_sofia, file sofia.c line 6515, I added the last OR condition. } else if (!strncasecmp(un->un_name, "X-", 2) || !strncasecmp(un->un_name, "P-", 2) || !strncasecmp(un->un_name, "Cisco", 5)) { and now I can retrieve this header from the dialplan just like this: ${sip_h_Cisco-Guid}"/> Thanks to Swk. -- View this message in context: http://freeswitch-users.2379917.n2.nabble.com/Sip-Headers-advice-Not-parsin… Sent from the freeswitch-users mailing list archive at Nabble.com.

5 5

ptime
by michael knop 02 Aug '11

02 Aug '11

Hi all, I’m using FreeSWITCH Version 1.0.head (git-0fc8050 2011-07-31 22-14-06 -0500) and I’m dealing with a problem <http://pastebin.freeswitch.org/16913> that sounds like the one that is described in the Jira bug report <http://jira.freeswitch.org/browse/FS-2003>. Is the patch to reset the timestamp detector every time FreeSWITCH gets a new sdp or reply to invite still active? Thanks, micha

1 1