May 2012 - FreeSWITCH-users - FreeSWITCH Mailing Lists

HA using redis for tracking calls status
by David Notivol 18 May '12

18 May '12

Hello, I'm thinking on going to a High-Availability scenario with FreeSwitch. Reading the wiki I see the solution for sharing the call status between the master/slave servers is an ODBC source, as Postgresql/MySQL. But I'm a bit concerned of how this could impact the performance of the machine. My idea was to have redis doing this function in this scenario. Do you think this is possible? Does anyone have any expierence with this? Any advice? Thanks a lot. -- Regards, David Notivol

2 1

transfer caller ID problem and plycom
by ghallab 18 May '12

18 May '12

When their Polycom 550 rings the user see the caller ID of the person who transferred the call, once they answer the phone the caller ID of the transferred caller is displayed. I already tried the ignore display updates set to faulse and true in the vars.xml file. But don't have effect!

3 6

T.38/spandsp_fax interop problem
by Marcin Gozdalik 18 May '12

18 May '12

Hi I'm trying to send fax from FreeSWITCH to PSTN using T.38. FS is connected to Broadsoft softswitch. The softswitch is connected to PSTN using SIP-trunk to external provider. Apparently there is no T.38 on Broadsoft and it is provided by the external provider (i.e. Broadsoft provides only T.38 pass-through). I can receive faxes destined at FS without a problem. I call rxfax and FS sends re-INVITE to Broadsoft and T.38 transmission begins. However, when I want to send fax, I can't get re-INVITE from Broadsoft with T.38. From the company that operates the Broadsoft softswitch I received a PCAP of working transmission with Linksys/SPA2102-5.2.10: SPA2102 issues re-INVITE itself when sending fax and T.38 is negotiated correctly. Unfortunately I can't get FS to issue the re-INVITE. I thought that a part of the problem is that FS receives from Broadsoft RTP-event Fax ANS (32) messages and it somehow interferes with fax-detecting routing in spandsp. Unfortunately, disabling telephone-event altogether didn't help - Broadsoft does not sent RTP-event packets but still FS does not re-INVITE. My sending script is as follows: var variables = "{"; variables += "loopback_bowout=false,"; variables += "ignore_early_media=true,"; variables += "origination_caller_id_number='" + callerid + "',"; variables += "origination_caller_id_name='" + callername + "',"; variables += "fax_ident='" + callerid + "',"; variables += "fax_header='" + callername + "',"; variables += "fax_email='" + email + "',"; variables += "fax_dest='" + dest + "',"; variables += "fax_tiff='" + tiff + "',"; variables += "fax_attempt='" + attempt + "',"; variables += "fax_max_attempts='" + max_attempts + "',"; variables += "fax_delay='" + delay + "',"; variables += "fax_enable_t38=true,"; variables += "fax_enable_t38_insist=true,"; variables += "fax_enable_t38_request=true,"; variables += "fax_disable_v17=false,"; variables += "fax_use_ecm=true,"; //variables += "fax_force_caller=true,"; variables += "fax_verbose=true"; variables += "}"; var s = new Session(variables + "loopback/"+dest+"/pbx/XML"); hook_run = 0; s.setHangupHook(hangup_hook); if (s.ready()) { console_log("DEBUG", "calling txfax\n"); s.execute("txfax", tiff); console_log("DEBUG", "returning from txfax" + "\n"); } else { console_log("INFO", "sending fax from " + callerid + " to " + dest + ": could not connect: " + s.cause + "(" + s.causecode + ")" + "\n"); return; } I do send the fax through loopback as I need it to establish rating session. I have experimented with various settings in fax_* variables. The only combination where anything happens (there is some T.30 negotiation in the logs) is when I set fax_enable_t38_request to false - then FS tries to negotiate T.30 transmission over G.711. I've never observed T.38 connection when sending fax, but it always ends with (20) Received no response to DCS or TCF. Usually the only thing there is in the logs is: 2012-05-16 17:27:19.824842 [DEBUG] mod_spandsp_fax.c:1357 Raw read codec activation Success L16 20000 2012-05-16 17:27:19.824842 [DEBUG] mod_spandsp_fax.c:1373 Raw write codec activation Success L16 2012-05-16 17:27:22.124834 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier up (-2) in state 18 2012-05-16 17:27:22.884832 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier down (-1) in state 18 2012-05-16 17:27:23.444834 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier up (-2) in state 18 2012-05-16 17:27:23.784833 [DEBUG] switch_rtp.c:3452 RTP RECV DTMF 2012-05-16 17:27:23.804833 [DEBUG] switch_rtp.c:3452 RTP RECV DTMF 2012-05-16 17:27:26.544834 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC sig nal status is Carrier down (-1) in state 18 2012-05-16 17:27:26.544834 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier up (-2) in state 18 2012-05-16 17:27:27.264833 [DEBUG] switch_rtp.c:3452 RTP RECV DTMF 2012-05-16 17:27:27.284873 [DEBUG] switch_rtp.c:3452 RTP RECV DTMF 2012-05-16 17:27:27.604841 [DEBUG] switch_rtp.c:3452 RTP RECV DTMF 2012-05- 16 17:27:28.884836 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Abort (-8) in state 18 2012-05-16 17:27:28.884836 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier down (-1) in state 18 2012-05-16 17:27:30.704933 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier up (-2) in state 18 2012-05-16 17:27:30.764833 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier down (-1) in state 18 2012-05-16 17:27:31.904833 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier up (-2) in state 18 2012-05-16 17:27:34.544835 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier down (-1) in state 18 2012-05-16 17:27:37.544837 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier up (-2) in state 18 2012-05-16 17:27:40.164835 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier down (-1) in state 18 2012-05-16 17:27:41.184835 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier up (-2) in state 18 2012-05-16 17:27:41.224837 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Abort (-8) in state 18 2012-05-16 17:27:41.224837 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier down (-1) in state 18 2012-05-16 17:27:43.184834 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier up (-2) in state 18 2012-05-16 17:27:44.824838 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier down (-1) in state 18 2012-05-16 17:27:45.664833 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier up (-2) in state 18 2012-05-16 17:27:45.664833 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier down (-1) in state 18 2012-05-16 17:27:45.724837 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier up (-2) in state 18 2012-05-16 17:27:45.724837 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier down (-1) in state 18 2012-05-16 17:27:45.784837 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier up (-2) in state 18 2012-05-16 17:27:45.804837 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 HDLC signal status is Carrier down (-1) in state 18 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 Status changing to 'The call dropped prematurely' 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:491 ============================================================================== 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:504 Fax processing not successful - result (49) The call dropped prematurely. 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:509 Remote station id: 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:510 Local station id: 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:511 Pages transferred: 0 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:513 Total fax pages: 0 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:514 Image resolution: 0x0 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:515 Transfer Rate: 14400 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:517 ECM status off 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:518 remote country: 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:519 remote vendor: 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:520 remote model: 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:522 ============================================================================== 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 Changing from state 18 to 32 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:286 FLOW T.30 Changing from phase T30_PHASE_A_CNG to T30_PHASE_CALL_FINISHED 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:286 FLOW FAX Set rx type 9 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:286 FLOW FAX FAX exchange complete 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:286 FLOW FAX Set tx type 9 2012-05-16 17:27:46.304847 [DEBUG] mod_spandsp_fax.c:286 FLOW FAX FAX exchange complete The same script transmits faxes correctly with Cisco gateway, but Cisco correctly re-INVITEs FS when FS is transmitting fax. PCAP dumps available, but I'd like to keep them private and not post on public pastebin. I'd appreciate any help on this issue. Best regards -- Marcin Gozdalik

1 1

bridge a call to registered user
by Miha 18 May '12

18 May '12

Hi, just one question. I have created two profiles (suggested by @Michael) as I was having same NAT issue. Profiles are basically the same only registration port is different (5060, 5070) and NAT profile have set: <param name="NDLB-force-rport" value="true"/>. In my dialplan I have made bridge like this: <action application="bridge" data="{origination_callee_id_name='${effective_caller_id_name}',origination_calee_id_number='${effective_caller_id_name}'}user/${destination_number}.fs_test(a)fs_test.fs2.blabla.com" /> freeswitch@default> show re [ registrations] freeswitch@default> show registrations reg_user,realm,token,url,expires,network_ip,network_port,network_proto,hostname 018108753.fs_test,fs_test.fs2.blabla.com,40804176-fb1f176@172.31.1.190,sofia/internal/sip:018108753.fs_test@xxx.xxx.xxx.xxx:1265,1337169317,xxx.xxx.xxx.xxx,1265,udp,localhost.localdomain 018108752.fs_test,fs_test.fs2.blabla.comi,873cfe40-1224f25@172.31.1.171,sofia/nat/sip:018108752.fs_test@xxx.xxx.xxx.xxx:2097,1337170569,xxx.xxx.xxx.xxx,1920,udp,localhost.localdomain 2 total. Why only phone work which is registered on NAT profile. IF I call users which is registered on internal profile I get user not registered? Thanks! Miha

2 4

183 and no ringback when send_silence_when_idle is non-zero
by Phil Quesinberry 18 May '12

18 May '12

Ok. more info on this. In default Proxy-Media mode: When making a call through the gateway for PSTN termination, FS sends 180 - Ringing to the calling endpoint when 'send_silence_when_idle' is set to 0 and ringback tone is heard. FS sends 183 and early media to the calling endpoint when 'send_silence_when_idle' is set to a non-zero value and no ringback tone is heard. In Bypass-Media mode: Making that same call through the gateway, FS sends 180 - Ringing to the calling endpoint regardless of the 'send_silence_when_idle' value and ringback tone is heard. Is this the intended behavior? - Phil _____________________________________________ From: Phil Quesinberry Sent: Thursday, May 17, 2012 1:46 PM To: 'freeswitch-users(a)lists.freeswitch.org' Subject: Re: Loud whilte noise during sleep Actually it's one of the latest 1.2.0 revisions - b653c21. I didn't realize that with the quick look I had at the code. Curious. that makes me wonder if the Aastra phones are generating the noise when they don't see any media. A user complained about hearing the noise when first logging into voicemail, before she is asked for her PIN. I hadn't noticed it before because I normally do that on speaker but on the handset, it is quite noticeable. Is there a way to use send_silence_when_idle without having it kill ringback when set to a non-zero value? ---------- Anthony Minessale Thu May 17 20:21:35 MSD 2012 Which revision of the code are you looking at? The sleep sends dummy frames unless you set that send_silence_when_idle var, if its not set its probably not sending anything. You can see in the code that unless that var is set it wont be generating anything.

1 0

mod_voicemail vmain from dialplan
by Spencer Thomason 17 May '12

17 May '12

Hello all, I'm working on a multi tenant setup where on one of the domains I need to have a different action for the vmain key in mod_voicemail. I cannot modify the vmain extension in the default context as this would affect all domains. Is it possible to set the vmain extension from the dialplan or is there another way to handle this? Essentially I need to get the caller the option of pressing a key to be connected to after hours support when they get voicemail. Thanks, Spencer

2 1

Play wav before bridge in IVR, possible?
by Nikolay Rogoshchenkov 17 May '12

17 May '12

Hello, This is part of IVR config, and I would like to play announcement before bridge to 123(a)12.3.4.56, is it possible to do? =============================================== [...] <entry action="menu-exec-app" digits="4" param="transfer 110 XML cisco_cme"/>. <entry action="menu-exec-app" digits="4" param="transfer 115 XML cisco_cme"/> <entry action="menu-exec-app" digits="3" param="bridge sofia/internal/123%12.3.4.56"/> <entry action="menu-exec-app" digits="0" param="transfer 0099 XML public"/> <entry action="menu-top" digits="9"/>  </menu> </include> =============================================== Thank you. -- Nikolay

2 4

Polycom auto answer on active call
by Spencer Thomason 17 May '12

17 May '12

Hello, I'm having a problem with Polycom endpoints where the phone auto answers a page even if another call is in progress. My dialplan looks like this: <extension name="extension-intercom"> <condition field="destination_number" expression="^\*80(\d{3,7})$"> <action application="set" data="dialed_extension=$1"/> <action application="export" data="sip_auto_answer=true"/> <action application="bridge" data="user/${dialed_extension}@${domain_name}"/> </condition> </extension> Does anyone have any suggestions for an elegant way to solve this? Thanks, Spencer

2 7

Re: [Freeswitch-users] Bypass media succeeds from extension to gateway but fails from extension to extension
by Phil Quesinberry 17 May '12

17 May '12

Nevermind! As it turns out, FS does send the correct external/NAT IP addresses as long as BOTH extensions are set for bypass media. Forgive me for ever doubting... Of course it still doesn't work for reasons discussed earlier, although I think that it should work as long as there's no symmetric NAT involved. I'm guessing session border controller(s) will be the next step if we want to remove the endpoint to endpoint traffic from our network. Any suggestions? OpenSIPS certainly looks promising. Thanks, - Phil -----Original Message----- From: Phil Quesinberry Sent: Friday, May 11, 2012 2:41 PM To: 'freeswitch-users(a)lists.freeswitch.org' Subject: Bypass media succeeds from extension to gateway but fails from extension to extension Ken, Thanks for taking the time to write that out. I understand that and that makes perfect sense, although the endpoints in this case are configured to report their external IP addresses, either through STUN or a static NAT IP entry. What still is unclear to me is why this failed when FS was able to successfully bridge the same extension directly to the PSTN gateway. ... >From the traffic I pastebinned, we can see that Endpoint A is sending its external WAN address info to FS, right before FS sends Endpoint As internal LAN address to Endpoint B, so I would think that FS should be passing Endpoint As WAN address along instead of its LAN address for media: ... ---------- Ken Rice Thu May 10 19:08:53 MSD 2012 The contact IP has nothing to do with where the media goes... That¹s entirely defined in the SDP... Consider this Endpoint A (192.168.100.100) -> NAT A -> FreeSWITCH (4.2.2.2) -> NAT B -> Endpoint B (192.168.100.200) Now lets assume that NAT A and NAT B are 2 separate nat gateways and that Endpoint A and Endpoint B are on 2 different physical LANs... Telling Endpoint A to talk directly to Endpoint B without proxying media will never work since the endpoints think they are on the same LAN. There is no mechanism there to allow for the redirection and automagic adjustments of ports etc so that they can talk directly to each other... Now lets change this slightly so that endpoint B is 192.168.200.200. Unless NAT A knows how to get to 192.168.200.0/24 (assuming class C sized block) and NAT B knows how to get to 192.168.100.0/24 they are both going to use their default routing which is to NAT the outgoing RTP, and forward it to the next hop... Again, RTP will not make it to other side in either direction... FreeSWITCH cant compensate due to a number of factors... Your Endpoints have to be smart enough to actually compensate for the NAT in this situation OR your NAT boxes have to compensate for it... The simple answer, don¹t use bypass media in this situation, the complex answer I wont get into here... Stop by IRC and ask around... There is a 3rd option here you might want to consider, contact consulting at freeswitch.org <http://lists.freeswitch.org/mailman/listinfo/freeswitch-users> for some professional help... This may not be specifially what you need to get going as I have no clue what your skill level happens to be, and you did say you are still learning. Good Luck!

1 0

Re: [Freeswitch-users] Detect answer on FXO device
by cristian re 17 May '12

17 May '12

The ATA has both "call answer supervision" and "polarity reversal" but they doesn't works worldwide but only with some carriers. Is there a way to delegate this job to Freeswitch? using voice activity detection? Cristian >Technically, answer detection is the job of the ATA. The problem is that it >is an inexact science. I would look in the GXW docs to see if they have a >section on 'answer supervision'. It might also be in a section along with >'disconnect supervision.' > >-MC

2 1