DTLS failed After Unhold - [ice] ICE role change

[bojjakaseenadhreddy]

rtpengine version the issue has been seen with

12.2.2.0+0~mr12.2.2.0

Used distribution and its version

RHEL 8.10 fedora (Ootpa)

Linux kernel version used

4.18.0-553.85.1.el8_10.x86_64

CPU architecture issue was seen on (see uname -m)

x86_64

Expected behaviour you didn't see

After retrieving the call from Hold, DTLS should be successful and video should start.

Leg A
WebRTC jsSIP (chrome) <-> kamailio + rtpengine (both are proxy) <-> SBC (interface 1)

Leg B
Genesys + SBC (Interface 2) <-> WebRTC jsSIP (chrome

Unexpected behaviour you saw

After retrieving the call from Hold DTSL stuck in connecting state and eventually changed to failed because of this there is no two way video (audio is working fine).
Note: In the SIP signaling both sending and receiving ports changed when call is retrieved from Hold

[ice] ICE role change, now controlling
[ice] ICE/STUN response with unknown transaction received

Steps to reproduce the problem

No response

Additional program output to the terminal or logs illustrating the issue

i am getting such logs multiple times during the call
DEBUG: [f3a28uk7cptqqrptl5e5/sui2k0kfri/2 port 34699]: [ice] ICE role change, now controlling
DEBUG: [f3a28uk7cptqqrptl5e5/sui2k0kfri/2 port 34699]: [ice] Recalculating all ICE pair priorities
DEBUG: [f3a28uk7cptqqrptl5e5/sui2k0kfri/2 port 34699]: [ice] Triggering check for tPzL78BBEvSeH9XS:d7514f7334d0770:1

Anything else?

Pls let me know if you need complete logs, I can share the pcap and rtpengine debug logs privately.

Thank you!

[rfuchs]

Try this on a supported version first (12.5 for closest LTS) to see if it's actually still an issue.

[bojjakaseenadhreddy]

Thanks @rfuchs for your prompt response.

FYI, this issue is happening intermittently with 12.2.2.0
Sure, let me test this with 12.5.