March 2010 - FreeSWITCH-users - FreeSWITCH Mailing Lists

Spontaneous reboots of Polycom and Linksys phones
by Wellie Chao 26 Mar '10

26 Mar '10

I am using Freeswitch and I have some phones that spontaneously reboot (often in the middle of a call). I did some searching around on the web, but didn't find any information that really applied to the problem I'm experiencing. I have a few dozen Polycom phones and a handful of Linksys phones of various models: Polycom IP301, IP320, IP331, IP430, IP501, IP601; also Linksys SPA901, SPA921, SPA941. I also have snom phones, but they don't reboot spontaneously. It's only the Polycom and Linksys phones that reboot spontaneously, and it seems to be correlated to call volume. Heavy usage of the Polycom and Linksys phones causes them to reboot more frequently (i.e. shorter duration between the reboots). For people making 50 calls a day, the Polycom phones spontaneously reboot every three days or so (and always on a call, it seems). For people making one or two calls a day, the Polycom will go months without a spontaneous reboot. This correlation between call volume and reboot frequency leads me to believe that there is some bug in the Polycom firmware, perhaps some of overflow condition that accumulates over time and eventually causes a crash (and subsequent reboot). I was wondering if anybody else experienced something similar, and whether anyone had developed a solution. One thing I am considering is to set up a script to reboot the Polycom phones in the middle of the night every day, which would hopefully prevent them from spontaneously rebooting during the day by preventing the phones from ever reaching the overflow condition (if that is indeed what causes the reboots). A nightly reboot would be a hackish way of fixing the problem, so I'd also love to hear if anybody has any other speculations on what might be causing the spontaneous reboots. We have phones in multiple buildings, and we've discovered that the spontaneous reboots happen on phones on different networks, so it's not specific to a particular LAN segment. For the most part, we are using the standard config files for the Polycom phones and for the Linksys phones.

5 9

call info display
by Ken Gillett 26 Mar '10

26 Mar '10

Sorry to ask what are perhaps very basic questions, but could someone please enlighten me on FreeSwitch's capabilities with regard to information for display on the recipient's extension. When a call is received, the extension really needs to display information about the origin and the intended recipient of the call. Normally for an incoming external call, the former is supplied by the 'incoming CLI' (often linked to locally stored name) and the latter probably by associating a name with each incoming line, so in this way the recipient can see at a glance who called whom (or what). Am I right in thinking that within SIP there is the provision to send information about the calling party (origin), so the called party (recipient) doesn't have to rely on CLI and matching numbers to locally stored directories? So, how does FS handle all the above? What does it pass to the recipient, for both external and internal calls? Is this part of FS configuration? IOW, do you have programmatic access to such information and can you specify exactly what is sent to an extension? Ken G i l l e t t _/_/_/_/_/_/_/_/

5 8

Guidance on debugging audio glitches
by Chris Fowler 25 Mar '10

25 Mar '10

Hi, Here's the problem I'm "hearing": Test Case #1 x1001 calls x1002 - Audio is rock solid Test Case #2 x1001 calls x3001 (Conference Bridge) x1002 calls x3001 (Conference Bridge) - Audio is mostly perfect. However, occasional "glitches" in the audio can be heard Notes: - Bypass media is switched off. FreeSWITCH routes audio for both the bridged, and the conferenced call - rtp-autoflush-during-bridge is "true" on Sofia profile internal.xml - I enabled the debug trace for auto-flush in switch_rtp.c (line 1895) and see occasional auto flush of rtp packets for *both* test cases. The glitches don't coincide with the flushing. In fact with flushing you can't hear when data gets flushed (2x20ms packets). Environment - CentOS 5.4 - Tested on AWS EC2 Small, Medium, Large instances. On Medium and Large CPU affinity for FreeSWITCH set to the 2nd core. Questions: - What's different between bridging a call and a conference call run through mod_conference in terms of audio processing? - Thoughts on how to debug this? Thanks, Chris.

2 1

Disabling mod_xml_rpc
by Joseph L. Casale 25 Mar '10

25 Mar '10

If the web interface is not needed, can this module be left unloaded or do other services depend on it? Thanks, jlc

3 3

Per User Music on Hold
by Matt Putnam 25 Mar '10

25 Mar '10

Is there a way to change music on hold per user instead of per channel? I have one user that I would like to have different hold music than the default. I can get it to work outbound by setting it on the channel but inbound I cant seem to find a way to change its always the default specified in vars.xml.

2 1

Media traffic to bypass FreeSWITCH and to be exchanged by peers only
by Vitalii Colosov 25 Mar '10

25 Mar '10

Hi, Very newbie question: Currently both SIP and RTP traffic passes thru the FreeSWITCH server. Client A -->SIP and RTP -->FS server-->Client B Is it possible that only SIP traffic will pass thru FS, and media will be exchanged between Client A and Client B directly? Client A -->SIP -->FS server-->Client B Client A -->RTP -->Client B How can I configure this? Sure both client A and B will use same codec, otherwise FS Server is a must for transcoding, right? Or is a must for anything else... Thank you, vIT

2 3

international prefix
by Madovsky 25 Mar '10

25 Mar '10

I try to manage trunks that allow different international prefix. for example, trunk A allows "011" and trunk B "00", and for domestic calls trunk B allows no prefix. it's a problem when I use mod distributor since I can't gues which trunk will be used for the bridge. I'm not sure that lcr is a good solution for this problem as it makes more DB overload. Any idea ? Thanks

3 3

send fake UDP packets
by Madovsky 25 Mar '10

25 Mar '10

I use an audio stream relayed by a client that has no audio DSP card. so when I stop to talk, no UDP audio packet is sent to FS. Do you know a trick to send fake UDP packet to keep the call alive ? (I need to keep rtp-timeout short to hang up leg B when leg A hangs up) Thanks F

3 6

Re: [Freeswitch-users] Guidance on debugging audio glitches
by Chris Fowler 25 Mar '10

25 Mar '10

Forgot to mention : Running trunk - build 17104 From: Chris Fowler Sent: Thursday, March 25, 2010 12:48 PM To: 'freeswitch-users(a)lists.freeswitch.org' Subject: Guidance on debugging audio glitches Hi, Here's the problem I'm "hearing": Test Case #1 x1001 calls x1002 - Audio is rock solid Test Case #2 x1001 calls x3001 (Conference Bridge) x1002 calls x3001 (Conference Bridge) - Audio is mostly perfect. However, occasional "glitches" in the audio can be heard Notes: - Bypass media is switched off. FreeSWITCH routes audio for both the bridged, and the conferenced call - rtp-autoflush-during-bridge is "true" on Sofia profile internal.xml - I enabled the debug trace for auto-flush in switch_rtp.c (line 1895) and see occasional auto flush of rtp packets for *both* test cases. The glitches don't coincide with the flushing. In fact with flushing you can't hear when data gets flushed (2x20ms packets). Environment - CentOS 5.4 - Tested on AWS EC2 Small, Medium, Large instances. On Medium and Large CPU affinity for FreeSWITCH set to the 2nd core. Questions: - What's different between bridging a call and a conference call run through mod_conference in terms of audio processing? - Thoughts on how to debug this? Thanks, Chris.

2 1

variables scope
by Madovsky 25 Mar '10

25 Mar '10

Hello, what is the scope of variables set in the gateways config ? also is there any wiki that explains the scope of different vars ? Thanks

2 2