- FreeSWITCH-users - FreeSWITCH Mailing Lists

FreeSWITCH 1.2.17 crash
by Hector Geraldino 11 Sep '14

11 Sep '14

Hi everyone! In the past few days we've experienced a few crashes of FreeSWITCH (1.2.17 64bit) running on a RHEL 6.4 (Carbon) server. I've collected the host info, gdb traces, etc. using the fscore_pb script included on the FS sources. Link to pastebin: http://pastebin.freeswitch.org/23266 I know this FreeSWITCH version is relatively old, but I can't really upgrade this server to the new 1.4 stable branch as it requires newer versions of some libs (sqlite, openssl, etc) that are not available for this RHEL version. I simply can't upgrade the server, and my hope is that some of you can provide some clues of what might be the root cause of the crashing. Thanks all for your time, Hector

5 7

media bug error in log
by Doron Kruh 11 Sep '14

11 Sep '14

Hi All, I am using v1.2.stable. In my scenario I am bridging 2 legs. I am recording each leg using record_session application in the dial plan. I am getting errors throughout the session like these: 2014-09-11 13:23:01.891833 [ERR] switch_core_media_bug.c:226 READ BUFFER 320 WRITE BUFFER 0 2014-09-11 13:23:01.891833 [ERR] switch_core_media_bug.c:226 READ BUFFER 0 WRITE BUFFER 0 I did not see those error before I updated to the latest stable branch (I was using the old git repository). Can some one please direct me to what am I doing wrong? Thanks, Doron. -- Doron Kruh, R&D Team Leader Lexifone , Hapalyam 7 Haifa. Mobile: 972-52-3823361 Office: 972-4-9126793

2 1

voicemail: always wait 30 seconds
by Micha Knop 11 Sep '14

11 Sep '14

Hi all! In the FreeSWITCH default configuration, a call to a registered extension is routed to voicemail after 30 seconds. If the extension is not registered, the call is routed immediately to voicemail. I would like to always wait 30 seconds before the call goes to voicemail. From the outside you should not be able to find out if an extension is registered or not. I have already experimented with the following setting: <action application = "set" data = "continue_on_fail = USER_NOT_REGISTERED" /> Unfortunately unsuccessful. Can someone give me a hint? /micha

3 4

SIP OPTIONS pinger
by Alex Lake 11 Sep '14

11 Sep '14

I want to get an idea of how well connected my fs boxes are to certain SIP servers that we use. The idea might be to send a SIP options packet to a remote address and time how long it took to respond. I suspect that fs can do something like this, but I was also wondering if there exists an off-the-shelf Linux package (a bit like mrt) that can do this. Failing that, some code in (say) PHP that can do it? I would have some kind of scheduled job that would send out a SIP options packet every 10s or so and keep track of the rtt.

3 4

does Javascript fetchURL cache results?
by Grant Bagdasarian 11 Sep '14

11 Sep '14

Hello, In one of our Javascript FS applications we're using the fetchUrl function, but it seems like it is always returning the same value. When accessing the url through the browser the results are different for each request, but when FS does the request the results are always the same. I even executed the curl [URL] command in linux and fs_cli, both returned different values every time. Is FS caching the results when using the fetchUrl function? The url isn't even accessed that often, probably a few times a day. Regards, Grant

2 2

mod_lcr and nibblebill
by FERNANDO VILLARROEL 10 Sep '14

10 Sep '14

Dear all. I am trying to use the both modules, but i have problems: Always the balance decreases and when i cancel the call the destination number ringing anyway. Also if the customer does not have balance, to try to make a call this goes smoothly in other words the calls always comes. The log http://pastebin.freeswitch.org/23267 My lcr.conf.xml <configuration name="lcr.conf" description="LCR Configuration"> <settings>  <param name="odbc-dsn" value="freeswitch:freeswitch:vmo123"/> </settings> <profiles> <profile name="default"> <param name="id" value="0"/> <param name="order_by" value="rate,quality,reliability"/> </profile> <profile name="qual_rel"> <param name="id" value="1"/> <param name="order_by" value="quality,reliability"/> </profile> <profile name="rel_qual"> <param name="id" value="2"/> <param name="order_by" value="reliability,quality"/> </profile> <profile name="pg_prefix2"> <param name="order_by" value="rate,quality,reliability"/> <param name="export_fields" value="nibble_account,nibble_rate,lowbal_amt,nobal_amt"/> <param name="custom_sql" value=" SELECT l.digits AS lcr_digits, c.id AS lcr_carrierid_field, c.carrier_name AS lcr_carrier_name, l.${lcr_rate_field} as lcr_rate_field, cg.prefix AS lcr_gw_prefix, cg.suffix AS lcr_gw_suffix, l.lead_strip AS lcr_lead_strip, l.trail_strip AS lcr_trail_strip, l.prefix AS lcr_prefix, l.suffix AS lcr_suffix, a.id AS nibble_account, l.${lcr_rate_field} AS nibble_rate, a.warn_limit AS lowbal_amt, a.nobal_limit AS nobal_amt, a.id_tarifa AS tarifa FROM lcr l JOIN carriers c ON l.carrier_id=c.id JOIN carrier_gateway cg ON c.id=cg.carrier_id JOIN auth_user a ON a.id = ${accountcode} JOIN tarifas t ON t.id = a.id_tarifa WHERE c.enabled = '1' AND cg.enabled = '1' AND l.enabled = '1' AND l.digits @> '${lcr_query_digits}' AND CURRENT_TIMESTAMP BETWEEN date_start AND date_end ORDER BY digits DESC, ${lcr_rate_field}, random(); "/> </profile> </profiles> </configuration> My Dialplan: <extension name="tests"> <condition field="destination_number" expression="^([23]\d{3})$"> <action application="lcr" data="$1 pg_prefix2"/> <action application="bridge" data="${lcr_auto_route}"/> </condition> </extension> What i doing wrong i am follow http://wiki.freeswitch.org/wiki/Mod_lcr#Integrating_with_mod_nibblebill I appreciate some help. regards

1 0

WebRTC meeting Norfolk, 15 October 2014
by Daniel Pocock 10 Sep '14

10 Sep '14

I'll be in Norfolk, VA for xTupleCon in October On 15 October, there will be two events for WebRTC: 14:15 a talk about the xTuple WebRTC extension at xTupleCon - must register for xTupleCon to attend this 17:30 a technical / developer workshop at xTuple's offices - free, anybody welcome, even if not attending xTupleCon, RSVP through Eventbrite[1] Please see my blog[2] for more comments about all of this and feel free to email me in advance if you have questions about it or if you may like to meet up there. 1. http://www.eventbrite.com/e/browser-based-webrtc-telephony-for-web-apps-wor… 2. http://danielpocock.com/xtuplecon-webrtc-talk-schedule-change

1 0

Channel is hungup and application 'curl' does not have the zombie_exec flag.
by Phillip Jones 10 Sep '14

10 Sep '14

Hi all, This a very simple custom VM that plays a greeting, bong and then records the callers message. After the caller has hung up, I need to call a webservice through curl. The zombie_exec flag gets me to the last line but the curl does not execute complaining: "Channel is hungup and application 'curl' does not have the zombie_exec flag." How do I apply the zombie flag to CURL? <condition field="${ForwardingNumber}" expression="([1-9]),(.*)" break="never"> <action application="set_zombie_exec"/> <action application="answer" /> <action application="sleep" data="2000"/> <action application="playback" data="$${user_vm_greeting}\$2"/> <action application="playback" data="tone_stream://$${vm_bong}"/> <action application="record" data="$${user_vm_message}/${accountno}-${uuid}.wav $${maxreclen} $${silencethreshold} $${silencehits}"/> <action application="curl" data=" http://localhost:58784/api/Voicemail/InsertVoicemail"/> </condition> Thanks in anticipation. Phil

1 0

NOTIFY's Cseq too high?
by Yehavi Bourvine 10 Sep '14

10 Sep '14

Hi, My Polycom phones stopped showing presence today. They reply with "loop detected" after they receive the NOTIFY message. In their Syslog they complain that Cseq is outside boundaries... The Cseq for INVITE is much smaller. How can I reset the Cseq to lower values? Thanks! __Yehavi:

5 13

Double NAT
by k 10 Sep '14

10 Sep '14

Hi, all ! I'm looking for update to the problem mentioned here : https://wiki.freeswitch.org/wiki/Auto_NAT ----------------- Doublenat with eyebeam/snom : Eyebeam/snom1 <> NAT <> Internet <> NAT <> FS Also Eyebeam/snom2 <> Internet <> NAT <> FS Both "work" with audio both ways. snom1 could call snom2 and vice versa and also other phones on the LAN. Found that call is hung up with a "BYE" from FS stating reason is "ACK timeout" although "ACK" was being received at FS (and there was still audio, so shouldn't be due to RTP timeouts). Not sure if this is due to some wrong setup etc. Will update once it is clear. Eyebeam had stun enabled. FS had stun disabled and autonat enabled. Router was netgear FVS318 wireless. Same issue with snom 320 as well. ----------------- I'm experiencing exact the same thing with the following scheme : eyebeam <> NAT <> FS <> NAT <> eyebeam stun and ice are enabled on eyebeam, FS is configured for inbound_bypass_media, TLS, SRTP works well until both clients behind nat. Experience random hangups in several seconds .. several minutes

1 0