October 2016 - FreeSWITCH-users - FreeSWITCH Mailing Lists

mod_python and new Session from freeswitch.Session
by Андрей Асташов 10 Oct '16

10 Oct '16

Hi Guys. >From Dialplan run scripts in Python, from which I perform freeswitch.session("......"). That is, create a new session. And if I do not get an answer for a long time from the gateway, the execution of other scripts in Python is delayed until such time as the session will not be installed. Please tell me how you can create a new session, and do not block completely mod_python. I understand that once had to be done through ESL, but now everything has already been written by the script from the dialplan. And it works, if it does not create a new session. freeswitch@Test-Environment> # # Execute python script from fs_cli(in dialer mode): # fs_cli --retry --batchmode --interrupt --execute="python python_test_script" # 2016-10-07 17:55:29.970991 [NOTICE] mod_python.c:212 Invoking py module: python_test_script 2016-10-07 17:55:30.119989 [NOTICE] switch_cpp.cpp:1360 -->(489d5441-701d-4a98-80ca-3389b00fa98a) Create a new session through freeswitch.Session("sofia/gateway/Goip_One/+71234567890@192.168.100.28:5060 ") 2016-10-07 17:55:30.119989 [NOTICE] switch_channel.c:1104 New Channel sofia/external-g-eth0.nostun.noauth/%2B71234567890@192.168.100.28:5060 [489d5441-701d-4a98-80ca-3389b00fa98a] freeswitch@Test-Environment> # # New incoming call # 2016-10-07 17:55:33.920114 [NOTICE] switch_channel.c:1104 New Channel sofia/external-g-eth0.stun.auth/384851(a)1.2.3.4 [e11d39b8-0479-4d4b-bd8b-28634f0a970a] 2016-10-07 17:55:33.920114 [INFO] mod_dialplan_xml.c:637 Processing 384851 <384851>->318386 in context hive-context # # Execute second python script from dialplan: # 2016-10-07 17:55:33.920114 [NOTICE] mod_python.c:212 Invoking py module: python_test_script freeswitch@Test-Environment> # # 12 seconds of silence during which any python script will not be executed and will be expected to complete # freeswitch@Test-Environment> 2016-10-07 17:55:45.140231 [NOTICE] sofia.c:8078 Hangup sofia/external-g-eth0.nostun.noauth/%2B71234567890@192.168.100.28:5060 [CS_CONSUME_MEDIA] [UNALLOCATED_NUMBER] 2016-10-07 17:55:45.160086 [NOTICE] switch_cpp.cpp:1360 -->(489d5441-701d-4a98-80ca-3389b00fa98a) After Create a new session through freeswitch.Session 2016-10-07 17:55:45.160086 [NOTICE] switch_core_session.c:1665 Session 15 (sofia/external-g-eth0.nostun.noauth/%2B71234567890@192.168.100.28:5060) Ended 2016-10-07 17:55:45.160086 [NOTICE] switch_core_session.c:1669 Close Channel sofia/external-g-eth0.nostun.noauth/% 2B71234567890@192.168.100.28:5060 [CS_DESTROY] # # Continue executing second python script(from 2016-10-07 17:55:33.920114) after any answer from first python new freeswitch.Session # 2016-10-07 17:55:45.160086 [ALERT] switch_cpp.cpp:1360 (mtid:3656): Traceback 280920161717(e11d39b8-0479-4d4b-bd8b-28634f0a970a): --> 12 <--; now: 1475841345; python_start_time:1475841333;

4 6

Using presence or state to avoid intercom barge-in
by Phil Quesinberry 09 Oct '16

09 Oct '16

Hi, We're running into an issue with Cisco SPA50x phones where an intercom call to an extension either places the currently active call on hold or hangs up on it and switches to the intercom call. We'd like to not have this happen if the destination extension is in use. The only way I know of to do this would involve reconfiguring all of the phones to share presence and then check it in the dialplan before bridging. Is there a convenient "no-barge" variable or similar that can be set so FS won't send the invite if the extension is in use or a way to check the extension's state from the dialplan? I'm guessing that's a NO but I had to ask! Alternatively, is there a way to get the Cisco phones not to interrupt the call in progress when a call comes in? Instead of setting sip_auto_answer=true, I've also tried setting sip_h_Call-Info=<sip:ip.address.of.server>;answer-after=0 which appears to stop this behavior on the Aastra phones but not the Cisco phones, although I'm not entirely sure about this as I now can't seem to reproduce the undesired behavior on the Aastra phones... Dialplan entry for the intercom as it stands is as follows: <extension name="extension-intercom-fix" continue="true"> <condition field="destination_number" expression="^\*8(\d{2,7})$" > <action application="set" data="dialed_extension=$1" /> <action application="bridge" data="user/${dialed_extension}@${domain_name}" /> <action application="export" data="sip_auto_answer=true" /> </condition> </extension> Any thoughts or advice would be appreciated. Thanks! Phil Quesinberry Q Systems Engineering, Inc. Embedded Systems, Telecom, IT (410) 969-8002 http://www.qsystemsengineering.com <http://www.qsystemsengineering.com/>

2 1

Freeswitch HA postgresql BDR mod_sofia fails to start
by Arsen 08 Oct '16

08 Oct '16

Hi guys, I am running two freeswitch instances in HA cluster in OVH cloud VPS. with postgresql in the core postgresql BDR master-master replication sometimes (almost always) when I start/restart freeswitch service or mod_sofia I get errors: 2016-10-07 19:22:22.499304 [CRIT] switch_pgsql.c:360 Query (create index sr_call_id on sip_registrations (call_id)) took too long to complete or database not responding. 2016-10-07 19:22:22.599306 [CRIT] switch_pgsql.c:360 Query (insert into tasks (task_id, task_desc, task_group, task_runtime, task_sql_manager, hostname) values(1,'zrtp_cache_save','core',1475869032,0,'fusion01-2')) took too long to complete or database not responding. 2016-10-07 19:22:22.599306 [DEBUG] switch_pgsql.c:415 Query (insert into tasks (task_id, task_desc, task_group, task_runtime, task_sql_manager, hostname) values(2,'heartbeat','core',1475868132,0,'fusion01-2')) returned PGRES_FATAL_ERROR 2016-10-07 19:22:22.599306 [ERR] switch_pgsql.c:656 Error executing query: ERROR: current transaction is aborted, commands ignored until end of transaction block ..and sofia can not initialize profiles or only one profile or freeswitch fails to start at all I can see some queries on freeswitch db with state=active when I execute (select * from pg_stat_activity) even when FS is down. top - 19:39:38 up 3:07, 7 users, load average: 1.03, 1.04, 0.95 Tasks: 126 total, 2 running, 124 sleeping, 0 stopped, 0 zombie %Cpu0 : 93.0 us, 0.3 sy, 0.0 ni, 6.6 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st %Cpu1 : 7.1 us, 0.7 sy, 0.0 ni, 92.3 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st Release: Debian 8.6 freeswitch 1.6.11~19~5e413fe-1~jessie+1 amd64 postgresql-bdr-9.4 9.4.9-1jessie FS config is default both VPS LAN interconnected OpenStack KVM 2 vCores 2.4 GHz 8 GB RAM 40 GB SSD Local RAID 10 any ideas ? -- Regards, Arsen.

1 2

User configuration
by Valter Nogueira 07 Oct '16

07 Oct '16

I would like to configure a user that would be allow to use only one codec and call only from a certain subnet like 192.168.5.x. Is that possible? Thank you Valter

3 4

SIP Proxy
by Valter Nogueira 07 Oct '16

07 Oct '16

In Asterisk, I have a "SIP Proxy", with 2 sip trunk. An external Asterisk dials in SIP-1 and asterisk, thru the dialplan, sends the call to SIP-2 (an ISTP) In Freeswitch, I don't know if I should use directory/user for "SIP-1" or if it is allowed to use a gateway inside internal sip_profile. Is it possible authenticate user by IP instead of user/password? Thanks, Valter

1 0

ESL terminate calls
by Gonzalo Gasca Meza 07 Oct '16

07 Oct '16

I'm launching API Calls using ESL in FS 1.6 (via python) For each new call, after call is hangup, I send a *uuid_kill* and then ESL.client.disconnect() Are both necessary, or I can skip *uuid_kill*? I'm trying to improve performance in my ESL and avoid unnecessary TCP connections to Freeswitch. Thanks -G

3 3

Asynchronous PTIME not supported
by Voxbox.io 07 Oct '16

07 Oct '16

Hi Guys, I'm getting an high CPU usage when managing calls. Looking at console I can see too many messages like this: [WARNING] switch_core_media.c:2042 Asynchronous PTIME not supported, changing our end from 20 to 10 Do you know what they means? It's a problem relating to reframing of packets? If so, can I reject calls with a different frame than I specify? Thanks Max

3 4

Kernel Hyperthreading ON or OFF?
by Raimund Sacherer 06 Oct '16

06 Oct '16

Hi, we have a new Freeswitch cluster with Dell R410. It has 2 CPUs (Xeon E5620, QuadCore. it's 8 cores withouth HT and 16 cores with HT. I googled with hyperthreading and freeswitch but some say turn it on it will give you a boost, others say turn it off, you will gain performance. Does anyone know if it's better to leave it on or turn it off? And some small explanation? I would be very gratefull for your help, best regards Raimund --

2 1

Performance testing
by Gonzalo Gasca Meza 06 Oct '16

06 Oct '16

Hi all, I'm testing Freeswitch using ESL interface, recording and SIP UDP Calls at a rate of around 5 CPS. Freeswitch receives ESL requests and proceeds to send call to PSTN Emulator, once call is answered I start recording. I have noticed around 40% CPU utilization when recording is not enabled and ~60%+ CPU when recording is. I created/mount a volume specifically to store the recording files. (ext4) *Server specifications:* EC2 server (t2.medium) 2 vCPUs, 2.5 GHz, Intel Xeon Family, 4 GiB memory. FreeSWITCH (Version 1.6.9 -16-d574870 64bit) I followed instructions at: https://wiki.freeswitch.org/wiki/Performance_testing_and_configurations *[Questions]* a) How to strip down Freeswitch and remove non-used modules? Features I need: *esl, recording, loopback, bridge, sip, spandsp, cdr_csv.* My modules: *fs_cli -x "show modules"* http://pastebin.com/BQSxzFDE b) I would like to see which module is using X/Y amount of CPU cycles/memory in Freeswitch is it possible?. *fs_cli -x 'show status'* UP 0 years, 0 days, 1 hour, 2 minutes, 15 seconds, 541 milliseconds, 286 microseconds FreeSWITCH (Version 1.6.9 -16-d574870 64bit) is ready 18192 session(s) since startup 90 session(s) - peak 93, last 5min 92 5 session(s) per Sec out of max 30, peak 7, last 5min 6 1000 session(s) max min idle cpu 0.00/75.57 Current Stack Size/Max 240K/8192K A sample image is below: 40% steady (no recording) a peak when I enabled recording. [image: Inline image 1] # top top - 03:25:16 up 22:42, 1 user, load average: 3.88, 5.74, 10.39 Tasks:* 68 *total,* 1 *running,* 67 *sleeping,* 0 *stopped,* 0 * zombie %Cpu(s):* 11.0 *us,* 2.3 *sy,* 0.0 *ni,* 85.7 *id,* 0.0 *wa,* 0.0 *hi,* 0.7 *si,* 0.4 *st KiB Mem: * 4057584 *total,* 865964 *used,* 3191620 *free,* 147088 * buffers KiB Swap:* 1048572 *total,* 0 *used,* 1048572 *free.* 213572 *cached Mem PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 26217 freeswi+ -2 -10 1427444 302584 16904 S *65.9* 7.5 28:47.49 freeswitch

2 4

uuid_broadcast att_xfer + ringback
by Igor Olhovskiy 06 Oct '16

06 Oct '16

Hi! I’m trying to make an att_xfer via uui_broadcast. Scenario is I have a call between internal (A) and external (B) profile. On both channels set variables ringback=local_stream://moh transfer_ringback=local_stream://moh hold_music=local_stream://moh I can confirm them via cli with uuid_getvar I’m trying to transfer call from A to C. (At the end - connect B and C) Than I’m giving command via event socket uuid_broadcast B_CALL_UUID attxfer::sofia/gateway/XXX/NUMBER bleg (Why I’m not giving uuid_broadcast to A_CALL_UUID - cause I’m already have B_CALL_UUID and finding out A_CALL_UUID is an additional step, that I want to avoid) But I don’t get ringback music in this channel, only silence. What I’m missing? -- Best regards, Igor

1 0