Cloud Native Broadcast Engineers - Software-Defined Broadcast Readiness

In a Dynamic Media Facility (DMF), software-defined live production runs as media functions on a container platform, a host platform, and infrastructure. Where SMPTE ST 2110 streams media over IP, the open-source Media eXchange Layer (MXL) exchanges it through shared memory inside a host and RDMA between hosts.

Most broadcast teams have never owned those lower layers, and DMF and MXL, both still new, don't specifically say how to assemble and run them. We see the work falling into these problem domains:

• Performance: on a shared Linux host, a media function can miss its frame because of how the kernel schedules it or what runs beside it, and RDMA between hosts needs a network tuned against packet loss. You get hosts configured for frame timing, with CPU pinning and NUMA placement set for each function, and a network specified for media exchange.
• Security and isolation: the container security that keeps vendors apart also removes privileges low latency needs, and MXL shares memory between vendors' functions. Each function runs from a verified image with only the privileges and flows it needs, on a zero trust network.
• Reliability and operability: containers start too slowly to recover inside a frame, and broadcast monitoring can't see what went wrong on the host. Standby capacity is ready to take over, so hosts can be patched without taking media workloads off air, and host-level telemetry shows why a frame was late.

LiveWyer has built and operated production container platforms since 2015 and supports the DMF and MXL initiatives. Our engineers work alongside your broadcast team and vendors as a partner. You can start with a workshop on one idea or infrastructure decision, then move at your pace to develop and prove solutions which best fit your business.

As you would expect, our engineers also have a deep understanding of the AWS technical stack and how solutions developed for cloud native broadcast are best aligned and tuned for the environment. From instance profiling to high performing network architecture we work with the best AWS has to offer such as EFA, RDMA and native PTP.