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

[Freeswitch-dev] ESL and the directory
by João Mesquita 22 Jul '09

22 Jul '09

I have been thinking about what would be the best direction to FSGui and I am leaning towards something like Asterisk Flash Operator ( http://www.asternic.org) and HUD (http://pbxtra.fonality.com/products/hud/) Since ESL has no comparison to Asterisk Manager whatsoever, it won't be too hard to get the application controlling all the extensions and doing all the cool CTI things such as record on demand, voicemail control and conference stuff, etc... Anyway, I was looking on how to achieve this goal and the first thing I would need is the domain list as well its users and groups. The ideal here would be having the directory section itself so I can parse it on my own and even add a few more variables to it so FSGui can control permissions and stuff like that. This would eliminate for good the need of a FSGui server to control this kinda of thing (that's what HUD does). I would like to open a discussion on how to achieve this since it will most definitely require some hacking on mod_commands since xml_locate is not able to provide it all as of now. So, here are the questions: 1. Using ESL for that is the best approach? If not, what would be? (I foresee the need of SSL on libesl soon ...) 2. Should I fetch root with xml_locate on ESL and parse whatever I am interested on FsGUI or should I add switch_xml_list_domains as well its API and whatever other methods are needed on xml retrieval? 1. switch_xml_locate_domain needs the domain name and at this point, I don't have it yet. 3. Do any of you see a problem with having additional vars on xml directory to control FsGUI permissions and options? If yes, what is the alternative solution? Anyway, I throwing this at the table so I can get input before making a decision. Hope this does not give you all too much trouble. Later, jmesquita

4 5

[Freeswitch-dev] Log->Line vs Log-Line in event socket header
by seven du 22 Jul '09

22 Jul '09

HI, I think it should be Log-Line, too simple to make a patch, but should I report on jira? version 14048. Dialplan: portaudio/1 Regex (FAIL) [group_dial_sales] destination_number(1) =~ /^2000$/ break=on-false Content-Type: log/data Content-Length: 75 Log-Level: 7 Text-Channel: 1 Log-File: mod_dialplan_xml.c Log-Func: dialplan_hunt Log->Line: 304 Dialplan: portaudio/1 parsing [default->group_dial_support] continue=false

4 3

[Freeswitch-dev] ClueCon 2009 - Check Out The 15" MacBook Pro That We're Giving Away
by Michael Collins 21 Jul '09

21 Jul '09

This just in: We have a picture of the incredible MacBook Pro that we will be giving away this year: http://cluecon.com/node/38 Remember, all paid attendees are eligible to win this beautiful unit, so register today! Call 877.742.CLUE and get signed up. -Michael

1 0

[Freeswitch-dev] ClueCon 2009 - GREAT NEWS - Hotel and early registration extended through July 27!
by Michael Collins 21 Jul '09

21 Jul '09

Hello everyone! We are happy to announce that the ClueCon team has been able to extend the early bird registration price AND the hotel room reservations through Monday July 27th. This is important because, outside of our ClueCon block of rooms, the Wyndham is completely filled up. Please make your hotel reservations right away. If you haven't secured your ClueCon registration yet then please call 877.742.CLUE and we will get you set up with the early bird special of $499. Don't delay! ClueCon 2009 starts two weeks from today. It will be here before you know it. See you all in Chicago! -Michael http://www.cluecon.com 877.742.CLUE

1 0

[Freeswitch-dev] ClueCon 2009 - Last Chance To Book At The Wyndham
by Michael Collins 20 Jul '09

20 Jul '09

Greetings! We would like to let everyone know that the hotel where ClueCon 2009 is being held - The Wyndham Chicago - is completely booked up outside of the block of rooms that we have reserved. There are still rooms available. However, after noon CST tomorrow, July 21, those rooms will no longer be reserved for ClueCon attendees so you have until then to get your reservations in. After noon tomorrow we will no longer be able to guarantee that you have an chance to register with the hotel, so please act fast! See you in Chicago. Thanks for supporting ClueCon! -Michael http://www.cluecon.com 877.742.CLUE

1 0

[Freeswitch-dev] modifying oz to support hairpin calls / bypass_media_after_bridge
by John Wehle 18 Jul '09

18 Jul '09

The particular application I have in mind is routing incoming data calls to the correct digital modem banks. Using T1 and zaptel I'd like to take a incoming call on channel X and cross connect it to channel Y when bridging. I believe that bypass_media_after_bridge is intended to control this behavior. On the zaptel side it appears this is done by issuing ZT_CHANCONFIG with the appropriate zt_chanconfig. Where I'm looking for guidance is what needs to be in mod_openzap for FreeSWITCH to activate this behavior. I.e.: a) at some point FreeSWITCH needs to indicate to the openzap endpoint that FS channel A should be bridged to FS channel B. How does this happen? b) openzap then needs to look at channel A and B to determine if they can be bridged at the openzap layer. How does it indicate to FreeSWITCH that the channels have been bridged by openzap? c) what does call teardown look like? -- John ------------------------------------------------------------------------- | Feith Systems | Voice: 1-215-646-8000 | Email: john(a)feith.com | | John Wehle | Fax: 1-215-540-5495 | | -------------------------------------------------------------------------

2 3

Re: [Freeswitch-dev] access session variables after hangup
by rentmycoder rentmycoder 16 Jul '09

16 Jul '09

Ok, let's explain some situations when cdr is not usable. Let's take a simple IVR, callers dial in, navigate in the menu system, at some point they hangup. My customer needs a realtime monitor which shows the active calls in the system each with realtime data (calldate,callerid,actual menuitem, actual action (playing a wav file or waiting for dtmf), total call time, calltime since actual menu entered, etc...). With asterisk i just use AGI to communicate with an external mysql database to store this information and an external website, which connects to this database and show;s it's realtime content in the administrator's browser. How can I do this with freeswitch most effectively? I could use event socket, but for this simple task I think lua would be more cost effective... But I cannot execute any script telling the database that the actual call is hanged up... any idea?

3 2

[Freeswitch-dev] ClueCon 2009 - News and Notes
by Michael Collins 16 Jul '09

16 Jul '09

Only 19 more days until ClueCon! If you haven't registered then please do so right away by calling 877.742.CLUE. This year's event is going to be awesome - you don't want to miss it! Some news and notes: We have two more sponsors for our event this year: Nokia and Twilio! We are happy to welcome them both to the conference this year. Many of our sponsors are supplying items for each attendee as well as some nice prizes that will be raffled off at the end of events on Thursday afternoon. We also have a very special guest who will be speaking at ClueCon: Philip Zimmermann! Philip is the author of PGP and co-author of ZRTP. We look forward to his presentation on Wednesday. Another security expert, Dan York of Voxeo, will also be speaking on Wednesday. All of those who are interested in security will want to pay special attention to Wednesday's program. Thanks for your support and we look forward to seeing you in a few weeks! -Michael Collins http://www.cluecon.com 877.742.CLUE

1 0

[Freeswitch-dev] Question about Valgrind output showing mem leak in FS
by freeswitch-dev＠lists.freeswitch.org 16 Jul '09

16 Jul '09

Hi, I run valgrind and get the following output. It points to some places where there are some "definitely memory leak". Is this a real issue or just a false alarm? [root@server ~]# cat /usr/local/freeswitch.bak/conf/test.log ==20667== Memcheck, a memory error detector. ==20667== Copyright (C) 2002-2006, and GNU GPL'd, by Julian Seward et al. ==20667== Using LibVEX rev 1658, a library for dynamic binary translation. ==20667== Copyright (C) 2004-2006, and GNU GPL'd, by OpenWorks LLP. ==20667== Using valgrind-3.2.1, a dynamic binary instrumentation framework. ==20667== Copyright (C) 2000-2006, and GNU GPL'd, by Julian Seward et al. ==20667== For more details, rerun with: -v ==20667== ==20667== My PID = 20667, parent PID = 19188. Prog and args are: ==20667== /usr/local/freeswitch.bak/bin/freeswitch ==20667== -vg ==20667== -log ==20667== /usr/local/freeswitch.bak/log/ ==20667== -conf ==20667== /usr/local/freeswitch.bak/conf/ ==20667== -db ==20667== /usr/local/freeswitch.bak/db/ ==20667== ==20667== Warning: set address range perms: large range 1024004096 (defined) ==20667== Warning: set address range perms: large range 1024004096 (defined) ==20667== Warning: set address range perms: large range 1024004096 (noaccess) ==20667== Warning: set address range perms: large range 1024004096 (noaccess) ==20667== Warning: set address range perms: large range 1024004096 (defined) ==20667== Warning: set address range perms: large range 1024004096 (defined) ==20667== Warning: set address range perms: large range 1024004096 (noaccess) ==20667== Warning: set address range perms: large range 1024004096 (noaccess) ==20667== ==20667== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 68 from 1) ==20667== malloc/free: in use at exit: 1,121,194 bytes in 426 blocks. ==20667== malloc/free: 4,603,109 allocs, 4,602,683 frees, 1,569,942,087 bytes allocated. ==20667== For counts of detected errors, rerun with: -v ==20667== searching for pointers to 426 not-freed blocks. ==20667== checked 2,477,072 bytes. ==20667== ==20667== 513 bytes in 3 blocks are definitely lost in loss record 15 of 33 ==20667== at 0x4A05809: malloc (vg_replace_malloc.c:149) ==20667== by 0x4D1AA25: crypto_alloc (alloc.c:97) ==20667== by 0x4D18551: aes_icm_alloc_ismacryp (aes_icm.c:115) ==20667== by 0x4D16F59: cipher_type_self_test (cipher.c:264) ==20667== by 0x4D1A456: crypto_kernel_load_cipher_type (crypto_kernel.c:310) ==20667== by 0x4D1A955: crypto_kernel_init (crypto_kernel.c:154) ==20667== by 0x4D14D68: srtp_init (srtp.c:1081) ==20667== by 0x4C83A1A: switch_rtp_init (switch_rtp.c:407) ==20667== by 0x4C70935: switch_core_init (switch_core.c:1112) ==20667== by 0x4C719AF: switch_core_init_and_modload (switch_core.c:1278) ==20667== by 0x4025BA: main (switch.c:701) ==20667== ==20667== ==20667== 576 bytes in 1 blocks are definitely lost in loss record 17 of 33 ==20667== at 0x4A05809: malloc (vg_replace_malloc.c:149) ==20667== by 0x64B80D2: sofia_glue_do_invite (sofia_glue.c:1488) ==20667== by 0x6490913: sofia_on_init (mod_sofia.c:102) ==20667== by 0x4C69523: switch_core_session_run (switch_core_state_machine.c:480) ==20667== by 0x4C66D0F: switch_core_session_thread (switch_core_session.c:1065) ==20667== by 0x3B85806366: start_thread (in /lib64/libpthread-2.5.so) ==20667== by 0x3B850D2F7C: clone (in /lib64/libc-2.5.so) ==20667== ==20667== ==20667== 640 bytes in 2 blocks are possibly lost in loss record 18 of 33 ==20667== at 0x4A04B32: calloc (vg_replace_malloc.c:279) ==20667== by 0x3B84C0FF12: _dl_allocate_tls (in /lib64/ld-2.5.so) ==20667== by 0x3B85806AC2: pthread_create@@GLIBC_2.2.5 (in /lib64/ libpthread-2.5.so) ==20667== by 0x4C6FB05: switch_core_launch_thread (switch_core.c:390) ==20667== by 0x4C77DE0: switch_loadable_module_init (switch_loadable_module.c:121) ==20667== by 0x4C71AA9: switch_core_init_and_modload (switch_core.c:1307) ==20667== by 0x4025BA: main (switch.c:701) ==20667== ==20667== ==20667== 1,152 bytes in 2 blocks are definitely lost in loss record 23 of 33 ==20667== at 0x4A05809: malloc (vg_replace_malloc.c:149) ==20667== by 0x6499E41: sofia_handle_sip_i_invite (sofia.c:4891) ==20667== by 0x649E733: sofia_event_callback (sofia.c:478) ==20667== by 0x6519D19: nua_application_event (nua_stack.c:393) ==20667== by 0x65971E8: su_base_port_execute_msgs (su_base_port.c:280) ==20667== by 0x6596F60: su_base_port_getmsgs (su_base_port.c:202) ==20667== by 0x6597503: su_base_port_step (su_base_port.c:473) ==20667== by 0x659A2C5: su_port_step (su_port.h:340) ==20667== by 0x659A28D: su_root_step (su_root.c:858) ==20667== by 0x64A8E9A: sofia_profile_thread_run (sofia.c:877) ==20667== by 0x3B85806366: start_thread (in /lib64/libpthread-2.5.so) ==20667== by 0x3B850D2F7C: clone (in /lib64/libc-2.5.so) ==20667== ==20667== ==20667== 2,211 bytes in 201 blocks are definitely lost in loss record 24 of 33 ==20667== at 0x4A05809: malloc (vg_replace_malloc.c:149) ==20667== by 0x3B85078AC1: strdup (in /lib64/libc-2.5.so) ==20667== by 0x72EE44F: ??? (mod_highroute.c:1445) ==20667== by 0x4C67197: switch_core_session_exec (switch_core_session.c:1473) ==20667== by 0x4C6761E: switch_core_session_execute_application (switch_core_session.c:1395) ==20667== by 0x4C6A892: switch_core_session_run (switch_core_state_machine.c:166) ==20667== by 0x4C66D0F: switch_core_session_thread (switch_core_session.c:1065) ==20667== by 0x3B85806366: start_thread (in /lib64/libpthread-2.5.so) ==20667== by 0x3B850D2F7C: clone (in /lib64/libc-2.5.so) ==20667== ==20667== LEAK SUMMARY: ==20667== definitely lost: 4,452 bytes in 207 blocks. ==20667== possibly lost: 640 bytes in 2 blocks. ==20667== still reachable: 1,116,102 bytes in 217 blocks. ==20667== suppressed: 0 bytes in 0 blocks. ==20667== Reachable blocks (those to which a pointer was found) are not shown. ==20667== To see them, rerun with: --show-reachable=yes

1 2

[Freeswitch-dev] Quick XML parser question
by Michael Collins 15 Jul '09

15 Jul '09

Guys, This dialplan entry does not cause an XML parser error: <extension name="XML Parse Test"> <condition field="destination_number" expression="^1234$"> <action application="answer"/>  <action application="log" data="INFO You see me" <action application="log" data="INFO You don't see me"/> <action application="log" data="INFO But you do see me"/> </condition> </extension> Just curious - is the parser supposed to bomb out when it sees two < characters without a > in between? -MC

1 0