as discussed many times already, there are several factors that affect audio quality in conferences or playback of audio files (regular bridged calls are less sensitive to this). 1) Less often than #2, but sometimes you don't get the clock that is precise enough. Conferencing and audio playback require a good reference clock, and some virtualized environments don't provide it. As noted somewhere in the wiki, by default AWS provides a kernel with really bad clock frequency, so you ought to change the kernel. 2) Any public VM hosting is a best-effort service, and you are very likely to have noisy neighbors. They all produce spikes in CPU load, and your VM is by no means prioritized against the others. So, your FS needs to compose an RTP frame and send it on-time, but it simply doesn't get the CPU cycles when it needs them. This leads to rather high jitter, and it's totally unpredictable. It may also depend on time of day or day of week, and you get totally different quality measurements with 5 minutes interval. So, for any kind of IVR or conferencing server, you ought to have physical hardware, or a hypervisor under your control, so that you allocate a guaranteed CPU resource to your FreeSWITCH machine. There's actually a quite broad choice of physical hardware hosting, so you can place your media servers on physical boxes, while leaving your application logic at your beloved AWS. The physical hardware doesn't have to be big: without transcoding, an average Celeron server can handle dozens of conference legs without problems. It's also quite easy to estimate your CPU capacity in advance, before sending production traffic: you create a conference, one leg goes to your ear, then few dozens of legs go to a silent sink on another server, and one leg is playing back some reference audio. Then you increase the number of participants until you start hearing quality impairments. There are also commercial tools for measuring voice quality, if you intend to run quality checks periodically. cheers, stanislav On Fri, Apr 14, 2017 at 6:48 PM, Bilal Dar <bilal@rgate-systems.com> wrote:
I have been struggling with an issue for almost 2 weeks.
Our regular calls have no quality issue and looking RTCP statistics network conditions are perfect. We have normally on peak hr 60 calls and around 10 conferences.
We have noticed that when we have 2 conferences of 4 or 5 participants, audio starts breaking for the users who are on conference. Regular calls do not experience any quality degradation.
I upgraded the server to specs of 30Gig memory and 8 vCPU but still the issue exists. Common thing I have noticed even during off-peak hrs is that two 4+ participant call can cause the issue.
I have ruled out network & hardware. Last change I made was moved all users to G.711 from G.722. Now I am not sure what other steps I can take. Appreciate any suggestions.
Thanks
_________________________________________________________________________ Professional FreeSWITCH Consulting Services: consulting@freeswitch.org http://www.freeswitchsolutions.com
Official FreeSWITCH Sites http://www.freeswitch.org http://confluence.freeswitch.org http://www.cluecon.com
FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org