Please loook in to these two log I am attaching with this.
one success without 408 , one with 408.
Key point we put outbound_proxy=sip:sipcore.us-east-1.uat.carestack:4060;lr which is a Round Robin DNS.
So when one request is failing through one proxy . Asterisk is actually sending second invite to another proxy. But in Debug logs the channel remain unchanged which leads to Ending the Already connected call.
Even through 408 happen for the initial invite through proxy1.
Please look into the trace and share your thoughts on my findings.
I also suspect that we are using outbound_proxy as wrong way.
Due to network firewall issue inital invite through one proxy willl not get any response from Device (No 100 or 180 response)
Asterisk Tries with second proxy got back (100 , 180 , followed by 200 okay)
During both time Asterisk dialing with same channel , so 408 response from proxy1 for initial invite is affecting already connected call through porxy2,
Checking PJSIP outbound_proxy can be used like this?.
This is not a blocker issue. but it will affect for our fail over implementation currently.
Hi, This issue was an never ending bug in asterisk side, which result in Ending already established customer call.
while the issue occurred because of some network issue at customer side, ending an established call had raised serious concern of reliability of our product.