How to Monitor Cloud-Based Low-Latency Streams at Scale

What is the essential monitoring checklist for maintaining efficiency, 弹性, and performance for large-scale low-latency streaming using cloud-based workflows? Allan McLennan, President, 2G Digital Post Optimization, discusses the key components and tools for an effective live stream monitoring strategy with Ryan Jespersen, Director of Product Strategy, 杜比.io，在这个剪辑中 流媒体 Connect.

McLennan asks Jespersen to break down the methods he uses to monitor streams at scale efficiently.

“It's very, very, very hard to scale anything these days,” Jespersen says. “我在。 Wowza when we built Wowza Cloud. 你见过 Mux come out of nowhere pretty much in the last five, six years to become one of the biggest providers in the industry. AWS元素 这个也是吗?. These are all cloud services that are black boxes for people who are using them. You send in an ingest and get an output, but you see nothing happening in the ecosystem. So what's really important when you build any cloud service is giving your customers access to what's happening in this black box.” He emphasizes that customer monitoring tools must include error detection, parameter monitoring, and client-side issue tracking elements.

Jespersen highlights that he was a part of a startup, Millicast, which 杜比 acquired. “That's our WebRTC streaming service. 我们两件事都做. We have an origin, and we have an edge. 我们的边缘, everything we do is stateful, meaning it's like the old days of RTMP, where it's very hard to scale because you're creating a peer-to-peer connection between an edge server and a client. We offer analytical tools for our customers to see and monitor if there are issues on the distribution side. So, we offer that through an API on the contribution and distribution side in our dashboard. And essentially, it's true of every cloud service.”

看 完整的会议 from 流媒体 Connect November 2023. We'll be back in person for 纽约流媒体 on May 20-22, 2024. 更多细节请点击这里.