By Tony Jones, Principal Technologist, MediaKind
A few years ago, broadcasters and content owners perceived the public cloud as an unsuitable option for replicating live TV quality or reliability. Today that has definitively changed.
Greatly helped by the emphasis given to professional media by the likes of Microsoft, Google, and AWS, the public cloud is now seen as eminently viable in most respects. These big cloud providers are becoming the system integrators of the 2020s. It is now up to the software vendors to provide the capabilities that seamlessly integrate into cloud infrastructures and deliver the tools that work the way customers need.
This is not as simple as it sounds. It is relatively easy to make something deployable as an application in a cloud environment. The tricky bit with media processing in the cloud is meeting the exacting expectations of a real-time broadcast TV service.
Unlike everyday internet applications, broadcast services are in the same failsafe category of availability as emergency services. For example, if an online ticketing application fails, it doesn’t cause any significant outage problem. If someone can’t make their booking, they can probably do it successfully five minutes later. But if you lose five minutes of a broadcast event, such as a high-value soccer match, that could be terminal for the business.
The expectations are leagues apart.
Also, most cloud-based applications are transactional. Online ticketing is precisely that. You make your request to book, the system transacts calling on data stored in the database, you get the response, and the transaction is complete. Live video is a different beast. By nature, live video is a continual flow of data that doesn’t want to be interrupted.
The other aspect that needs consideration is how to get real-time video content into the public cloud. Secure Reliable Transport (SRT) is an industry-standard solution that can achieve this by providing a robust stream connection with controlled end-to-end latency and encryption, enabling a new generation of real-time connections into and out of the public cloud. You can read more about SRT in my interview with the SRT Alliance last year.
The challenge in building a cloud-deployed broadcast-capable system is to meet those expectations within an IT environment that wasn’t initially designed for either five-nines availability or real-time content flow.
MediaKind has been targeting this challenge for over five years. It has taken us toward building our technology in a Kubernetes environment because that can be deployed just as easily into Google Cloud or Azure and AWS as it can on-prem. You can take the same code that can be cloud-deployed and build it into a physical appliance. Since it is all software, you can integrate a complete stack across the whole spectrum of different deployment models, but it operates the same way regardless.
The advantages are legion. Apart from giving operators a variety of deployment options – whether on-prem or in any form of public cloud – they can scale operations with cost efficiencies unachievable if attempted conventionally.
A number of our major tier 1 customers began moving in this direction some time ago. They wanted to eliminate manual operations and costly errors and outages by automating the process in the cloud. Not only are operations far less error-prone, but they are reproducible too. With software built on Kubernetes, the same results will be repeatable ad infinitum, a predictable model that is highly expensive – if not next to impossible – to achieve manually on-premises.
All of our media processing software is now available in a cloud-agnostic containerized deployable format. Within Cygnus (contribution and distribution solutions) and Aquila (consumer delivery solutions), we can address specific media industry and content delivery applications and use cases today.
MediaKind cloud technology can run on-prem or in the public cloud. It’s a cluster that works the same whether built on virtual machines in a private data center, in a headend, or the public cloud using third party servers and infrastructure.
These enhanced solutions provide an immediate response to the anticipated changes in professional media contribution, enabling broadcasters, operators, and service providers to transition to all-IP and the cloud, integrate existing and future codecs and standards, and embrace new flexible business models.
Cloud infrastructure is accepted as a viable option for broadcast today. Now, broadcasters and content owners can invest with confidence in media software, safe in the knowledge they will always remain current in a world of shifting media contribution technologies and infrastructure.