Not sending voice to Google Dialogflow?

[voipterminator]

While testing the exemplary node.js app with Dialogflow integration I've noticed the Freeswitch modules stopped working properly with DialogFlow in the recent versions.

Freeswitch 1.8 with modules built based on the rev 6e7ead3 worked fine but when building fresh system based on the Ansible role https://github.com/davehorton/ansible-role-fsmrf the system is not forking the voice to Google APIs and immediately reverting to the failure intent. This keeps repeating all the time.

Any advice where the problem might be?

I've attached the debug log from Freeswitch.
freeswitch_log.txt

[davehorton]

Actually, the latest example app, along with instructions for building a freeswitch that supports it (and by the way, https://github.com/davehorton/ansible-role-fsmrf now builds freeswitch 1.10.1 instead of 1.8) can be found here:

https://github.com/davehorton/drachtio-dialogflow-phone-gateway

Try that and let me know if you still have problems

[voipterminator]

I've just tried the latest demo app, the result is very similar but this time it just keeps repeating the "noinput" intent after the timeout configured in the json configuration file.
Log attached.
new_demo_app.txt

[davehorton]

Are you sure audio is actually arriving inbound to freeswitch?

You might either try capturing a network trace to verify that audio is coming in, or try running the call-recording-with-per-did-config branch, which actually lets you optionally record the audio streams. It would be good to verify that freeswitch is seeing the audio

[voipterminator]

You're right! I was testing in my local network the new sample app and didn't notice Freeswitch pushed public IP in the SDP.
I do confirm the new sample app works with the updated FS modules when installed via Ansible role.
Thanks for your help again!