Choppy Audio on 1 Instance

[Anaphylaxis]

We use AWS and GCP to host our instances. We have dozens of instances with absolutely no issues, however there is 1 instance that is having call quality issues in the form of choppy audio. Short legs such as "One, two, three" will sound ok, however if a word is drawn out, the audio issues are clear: "Thr...e...e...e....e" for example.
It's evident it's a setting in 3CX, as we backed up the instance and migrated from AWS to GCP, and the issue still persists. This happens for random calls, but also for *777 which leads me to believe it's not the SIP uplink either. The interesting thing is some calls are fine.
Has anyone seen anything like this before?

 

[AWS2P]

What kind of instance do you run for this pbx? a small one ?
is the pbx a big install or a small one?

Is your trunk sip provider a supported one ?
Codecs audio are set accordingly to what expect your sip provider ?

 

[Anaphylaxis]

What kind of instance do you run for this pbx? a small one ?
is the pbx a big install or a small one?

Is your trunk sip provider a supported one ?
Codecs audio are set accordingly to what expect your sip provider ?

t3a.small on aws, g1-small on GCP. Same as we run for every instance, no issues so far..
Relatively small compared to our other instances.
Yep.
Yes, they are.

 

[leejor]

It's evident it's a setting in 3CX, as we backed up the instance and migrated from AWS to GCP, and the issue still persists.

It may very well be a provider/network/ISP issue, not a 3CX issue. Changing the hosted location doesn't prove that it is the PBX (only that it's probably not AWS), when everything else remains the same. How are the sets configured at that location...STUN, behind an SBC? If it is not using an SBC, then all audio will probably being going back to the PBX rather than set to set directly. Have you done latency testing on the local network and the ISP? I assume that this happens extension to extension and on outside calls as well?

 

[Anaphylaxis]

It may very well be a provider/network/ISP issue, not a 3CX issue. Changing the hosted location doesn't prove that it is the PBX (only that it's probably not AWS), when everything else remains the same. How are the sets configured at that location...STUN, behind an SBC? If it is not using an SBC, then all audio will probably being going back to the PBX rather than set to set directly. Have you done latency testing on the local network and the ISP? I assume that this happens extension to extension and on outside calls as well?

This happens at multiple sites, and sites that also work with other instances just fine.
They are direct connections, no SBC or STUN.
Latency and speed are fine at every site. And no, your assumptions are wrong: *777 is not good however some outside calls are good (for example to my Xfinity Mobile cell phone), however we have had complaints of it not working for Verizon cell phones, so...

 

[leejor]

This happens at multiple sites, and sites that also work with other instances just fine.

In the first post, you stated...

We have dozens of instances with absolutely no issues, however there is 1 instance that is having call quality issues in the form of choppy audio.

So, for clarification, is it one site having issues, or multiple sites having issues?

 

[Anaphylaxis]

In the first post, you stated...



So, for clarification, is it one site having issues, or multiple sites having issues?

Every physical location (site) on this instance.