Media Push Stop API issue in Cluster mode

[yashtandon113]

Their is one AMS cluster running with two origin nodes.

When we call the start media push API using the domain that is mapped to load balancer, it starts the Chrome browser/driver session on any of the instances in the backend (for example, AMS Node A) and then publishes the WebRTC stream back to the server.

Now when the media push stream is created, it is created on any of the AMS nodes (for example, on AMS Node B) in the cluster. At that time, sometimes their is a possibility that the Chrome driver session running on node A and streamId created on node B.

Now if the stop API request goes to Node A, where the actual Chromedriver session is running, the media push stream will be stopped, but if the request goes to Node B, it causes the error:

{"success":false,"message":"Driver does not exist for stream id: mediapush","dataId":"","errorId":0}%