Medianet is a framework to simplify video and the operation of networks where video is running. One of the common ways to determine video behavior on a network is to simply look at devices and try to determine the content. Given Medianet’s architecture, they leverage the endpoint itself to understand video traffic, said Grivaud.
They are able to gather this information thanks to an application stack called MSI that sits on the endpoint itself. MSI’s job is to provide simplified autoconfiguration with the network. As part of that process, MSI can communicate to the network the characteristic of the communication. MSI is producing the metadata that feeds the network’s metadata infrastructure. Because this information is travelling on the network, and is not just being analyzed by a single device on the network, then anything on the network, no matter at what point-a switch, a router, whatever-can understand what the video data is and provide a treatment for that, said Grivaud.
That overall understanding of video behavior across a network manifests itself in traffic monitoring and quality of service (QoS) delivery, said Grivaud. With all this ongoing metadata you can easily answer the question, “What is the performance at any node for a given telepresence or video stream?” at any time.
Additionally, Cisco has introduced Mediatrace which allows you to follow a single video stream, usually a bad video stream, through every single point on the network. If you’re trying to isolate issues, it’s far more efficient to look at a single stream rather than all the traffic on the network. Before, they had no data to analyze a specific problem. People would complain they had a bad video experience. Since there was no way to capture that specific video stream there was nothing they could do. They simply didn’t have any information to analyze the problem. With Mediatrace you now have the information where you can determine the root cause of the problem, and you can isolate faults, said Grivaud.