Problems with calls are heard robotic

[liigi]

Hello everyone this is my cluster

VERSION: 2.14-763a
BUILD: 200719-1645
Version: 2.14b0.5
SVN Version: 3271
DB Schema Version: 1602
DB Schema Update Date: 2020-08-01 22:09:44


DIALER 1
13.34.0-vici

DIALER2
13.21.1-vici

I have a problem where certain calls are heard like a robot (like when there is lag, packet loss etc...) but this happens with some calls

I have noticed that these calls with problems occur when the CALL SERVER IP is different from the SERVER IP through which the call is being made

Could you help me diagnose what could be happening, it's okay if the call comes from another dialer

[jamiemurray]

This would suggest there is either an issue with one of the servers (perhaps overload) or there is a bad network connection.
Have you identified if the robotic calls are in both directions or just one of the following:

Agent on 1 & Call on 2
Call on 1 & Agent on 2

Also, describe your setup a bit more, are any of the dialers also the DB?
What's the hardware specs of both machines and install methods?
Are both servers located in the same place and on the same LAN?
How many clients (devices) are on that same LAN?

[okli]

Try enabling jitter buffers in iax.conf on each telephony server and do iax2 reload once enabled on each
Compare iax.conf samples for asterisk 1.4 included in vicidial code to get the options needed - \trunk\docs\conf_examples\iax.conf.sample-1.4
asterisk 11+ iax.conf sample do not have jitter buffer enabled.

We recently had the exact same issue where cross server IAX calls were affected the same way as you describe, even though servers are connected to the same switches and issue was there even with one agent. There was no jitter with SIP legs of these calls.
Similar setup in a other DC did not have any jitter on cross server IAX calls and I never had the time to get to the bottom if it is switch configuration like QoS, high switch utilisation for other traffic or IAX being temperamental.
At the time of our tests iax2 show stats was showing up to 30-40ms jitter while latencies between the servers using ping were steady 0.2-0.5 ms or so.

Hope this helps.

[liigi]

Hello, thanks for the answer.

i have only 2 dialers , they are connect in the same switche /subnet . I have checked and it seems to be fine (in terms of memory and cpu usage) 10% load ,

Boht have quad core CPU and 8 gb of RAM

[liigi]

Ok i will try that in iax.conf , thanks

[liigi]

[quote="okli"]


This is my iax.conf


[general]
bindport=4569 ; bindport and bindaddr may be specified
bindaddr=0.0.0.0 ; more than once to bind to multiple
;iaxcompat=yes
;nochecksums=no
delayreject=yes
;amaflags=default
adsi=no
;accountcode=vicidial
language=en
mohinterpret=passthrough
mohsuggest=default
disallow=all
allow=ulaw
allow=g729
allow=gsm
jitterbuffer=yes
forcejitterbuffer=no
maxjitterbuffer=1000
maxjitterinterps=10
resyncthreshold=1000
trunkfreq=20 ; How frequently to send trunk msgs (in ms)
trunktimestamps=yes
minregexpire = 60
maxregexpire = 360
iaxthreadcount = 50
iaxmaxthreadcount = 250

[liigi]

I see some error :

[Mar 24 14:13:49] WARNING[53480][C-0002021d]: func_hangupcause.c:140 hangupcause_read: Unable to find information for channel
[Mar 24 14:13:49] WARNING[53508][C-00020211]: func_hangupcause.c:140 hangupcause_read: Unable to find information for channel
[Mar 24 14:13:50] WARNING[53378][C-000201f8]: func_hangupcause.c:149 hangupcause_read: Information type not recognized (IAX2/VICIDAST2-7613,tech)

[Mar 24 14:14:09] WARNING[21174]: chan_iax2.c:1241 jb_warning_output: Resyncing the jb. last_delay 30675, this delay -27996, threshold 34480, new offset 35888


also see full lag /jitter

IAX2/127.0.0.1:41569 127.0.0.1 ASTblind 00694/09624 00092/00024 01000ms 0000ms 0056ms ulaw Tx:NEW Rx:ACK
(None) 172.16.20.41 (None) 01168/00000 00000/00000 00000ms 0000ms 0040ms Unknow Rx:Unknown Rx:Unknown
(None) 172.16.20.41 (None) 01672/00000 00000/00000 00000ms 0000ms 0040ms Unknow Rx:Unknown Rx:Unknown
(None) 172.16.20.41 (None) 02205/00000 00000/00000 00000ms 0000ms 0040ms Unknow Rx:Unknown Rx:Unknown
IAX2/VICIAST1-2909 172.16.20.41 VICIAST1 02909/07234 00025/00153 00049ms 0940ms 40140ms ulaw Rx:NEW Tx:ACK
IAX2/VICIAST1-3226 172.16.20.41 VICIAST1 03226/10886 00028/00025 00045ms 0001ms 0050ms ulaw Rx:NEW Tx:ACK
(None) 127.0.0.1 (None) 08009/00000 00000/00000 00000ms 0000ms 0040ms Unknow Rx:Unknown Rx:Unknown
IAX2/VICIAST1-8069 172.16.20.41 VICIAST1 08069/08272 00018/00015 00044ms 0001ms 0054ms ulaw Rx:NEW Tx:ACK
IAX2/ASTblind-9624 127.0.0.1 ASTblind 09624/00694 00024/00092 00056ms 14080ms 1000ms ulaw Rx:NEW Tx:ACK

[liigi]

HI
Any idea of ​​what to validate, why there is latency in the iax communication

[eccomecalderon]

Hi, I have the same problem

"I have noticed that these calls with problems occur when the CALL SERVER IP is different from the SERVER IP through which the call is being made"

Anybody know how to fix this or how stop balancing or stop this "jumps" between asterisk servers.???