Not so sure I agree with your call failure analysis with the ladder diagram at the end, and it seems you're not so sure either ;) You say that the reason you sent the CANCEL was because you never received the second 200, but the second 200 isn't technically late, regardless of whether it's for the PRACK or the INVITE. Nor is the first actually. Thoughts? Correct me if I'm wrong :)
@rnbrady2 The state machine of the call generator has a 500 ms timer to get the SDP information when configured for UDP "Early media" when you don't configure an 'optional' delay. The SBC sends the SDP in the 2nd 200 message, The moment it happened I was thinking state machine, IETF, and the configuration.. simultaneously. It's absolutely slowing down under load, but I could have made some adjustments to the configuration to be more 'forgiving' but I wasn't trying to emulate a polycom.
Not so sure I agree with your call failure analysis with the ladder diagram at the end, and it seems you're not so sure either ;) You say that the reason you sent the CANCEL was because you never received the second 200, but the second 200 isn't technically late, regardless of whether it's for the PRACK or the INVITE. Nor is the first actually. Thoughts? Correct me if I'm wrong :)
rnbrady2 7 months ago
@rnbrady2 The state machine of the call generator has a 500 ms timer to get the SDP information when configured for UDP "Early media" when you don't configure an 'optional' delay. The SBC sends the SDP in the 2nd 200 message, The moment it happened I was thinking state machine, IETF, and the configuration.. simultaneously. It's absolutely slowing down under load, but I could have made some adjustments to the configuration to be more 'forgiving' but I wasn't trying to emulate a polycom.
talk57 7 months ago