Bitmovin Introduces Managed On-Prem Encoding Service
The flexible new service runs on the same software stack as Bitmovin's cloud encoding service
Learn more about the companies mentioned in this article in the Sourcebook:
Earlier this week, Bitmovin launched a new managed on-premise encoding service based on Docker and Kubernetes, the same software stack the company uses for its cloud encoding service. We spoke with CEO and co-founder Stefan Lederer to find out more.
Streaming Media: Tell us about the new service.
Lederer: Basically, our new managed on-premise encoding service allows customers to run what's essentially the Bitmovin cloud encoding platform in their own data center or virtual private cloud. Technically, it's a service deployed on Kubernetes, which is an open source cluster manager originally designed by Google and donated to the Cloud Native Computing Foundation. This allows multiple disparate computers in a data center or cloud to function as a (more or less) single platform. The encoding and related services are actually performed by Docker containers that run on Kubernetes.
Streaming Media: Sounds like virtual machines. How is this different?
Lederer: There are a number of technical reasons that the containerized approach enabled by Kubernetes and Docker is superior, but the bottom line is that many large companies are moving towards Kubernetes-based clusters in their data centers or private clouds. They're using these clusters to run any number of applications, from accounting to databases. As far as we know, we're the first encoding platform that runs on Kubernetes and Docker, which makes us the only solution for enterprises running this platform.
Streaming Media: What are the advantages of this approach?
Lederer: It's a way for multiple enterprise applications to efficiently share a cluster of hardware resources. That is, Kubernetes can manage multiple requests for CPU resources from multiple applications and distribute these tasks over the instances in the cluster. It can even add and remove instances when necessary, which helps minimize operating expenses in the cloud or data center. When you don't need resources for encoding, it frees up resources for other enterprise-related tasks. If you need to spin up additional resources for time-critical tasks, you can do so.
Streaming Media: How does the new service work?
Lederer: From a deployment perspective, we deploy, monitor, upgrade, and support the system as a software-as-a-service (SaaS) that's running on your hardware or in your virtual private cloud. From a user perspective, it looks identical to our cloud service, and uses the same API.
Bitmovin's managed on-premise encoding runs on Kubernetes and Docker locally or in a virtual private cloud.
Streaming Media: Who's the target user?
Lederer: Obviously, you have to be running Kubernetes to be interested, and that's primarily large enterprises today. So that's the starting point.
Streaming Media: What can you tell me about pricing?
Lederer: In a data center you'll pay by the CPU. In a virtual private cloud, you'll pay based upon actual throughput, using a schema similar to our cloud SaaS service.