MicroServices Bring New Agility And Cost Savings To IP-Based Content Delivery
Avid’s all-software MediaCentral platform helps users create and deliver content across multiple channels by launching a dedicated app.
Broadcast equipment suppliers continue to add new types of IT-centric functionality to their signal distribution and playout platforms, leveraging the cost and operational benefits of COTS hardware and Software Defined Networks (SDN) as a way to attract new types of customers that increasingly require the handing of hundreds or thousands of media streams simultaneously and are looking to for a cost-effective way to do it.
In today’s competitive world, content (media files) has to be distributed to linear TV channels and internet streams quickly while simultaneously being converted into various encoded formats that target the wide spectrum of mobile devices that consumers are increasingly migrating to as their primary video viewing platforms. With sometimes thousands of on-demand sessions for the same content coming in at the same time, this needs to be a highly automated and seamless (to the content distributor and the end user) process in order to be successful.
IT-Centric Signal-Processing Environments On The Horizon
For the past three years the answer thus far has been the growing infiltration of Internet Protocol (IP) technology into the Broadcast plant that has enabled distributors to deploy cloud-based infrastructures, leverage SDNs and establish the virtualization of traditional hardware. These cloud-based services—that is, the handing and processing of signals at a remotely based facility to reduce OpEX and human resource costs—allow customers to scale up or down their hardware and software processing requirements (e.g., transcoding, storage and automated playout), depending upon the need at hand, without incurring significant CapEX costs.
The latest piece to the IT-in-Broadcast story is the introduction of what are being called “microservices,” which allow a customer to engage the same cloud-based signal processing and delivery services but in a much shorter time period. It’s meant that remote services don’t have to be procured for an entire month or week when only a few hours are needed. Spread that concept across an entire network with many channels and one can see the potential cost savings.
The vision is for a “Master Control” operator of the future to use a Web-based GUI that features a number of virtual buttons and pre-established templates to launch applications dedicated to bandwidth allocation, bit rates, signal processing, storage, digital ad insertion, internal distribution and playout applications. They can then use sub-applications (microServices) to reserve time-of-day and duration. Indeed, it’s supposed to be self-service, as easy as using a mobile phone or tablet interface.
Large portfolio companies like Avid (MediaCentral), Evertz Microsystems (evEDGE), Grass Valley (Stratus), Imagine Communications (Zenuim) and Sony (Media Backbone) are now all offering some type of an all-software IP-based signal processing platform approach to their customers and grabbing the attention, if not yet the checkbooks, of trade show attendees around the world. They are being joined by a series of smaller established and new companies to the space that all see the future in deploying traditional video processing and playout functionality as easily as opening an application on a mobile phone.
The micro services concept has attracted many large media organizations at various trade shows because they help get the most efficiency and economy out of networked production and distribution environments.
What Is A Microservice?
The answer to this question is different, depending upon whom you ask. Every company has their own way of deploying virtualized workflows in the most affordable and efficient way. Brick Eksten, Chief Product Officer at Imagine Communications (and a man who holds several patents on how to deploy microservices for use in the Broadcast industry), said that it’s not a virtualization of a piece of hardware.
“Microservices are about getting the most efficiency and economy out of a network by building against open standards and putting the absolute minimal function possible that delivers you the business value that you are seeking,” he said. “If you go beyond that, you are building a macro service.”
Eksten said there are two “ideals” in microservices, one is to put the minimum amount of functionality into the network that returns real business value and the other is to provide abstraction from the underlying system such that all communications is against a native IP and services fabric.
“The test is that if your microservice has one more function than it actually needs, it is not a microservice,” he said. “If it cannot operate against an all IP fabric and requires gateways, it is not a microservice. There are a few companies out there that are calling what they have microservices but in reality you have to go IP into a gateway device which is all baseband internally, uses programmable logic, and then goes through the opposite translation to get back to IP. That is not a microservice. A microservice should therefore be portable, it if it is not, if it requires specific topologies of applications it is probably just some form of plugin.”
All of Imagine’s products, including servers, transcoders, IP gateways, are now available as microservices through its Zenium “cloud-native” platform. Eksten said his company has spent years building a platform for the development, design, deployment and management of microservices because microservices support a number of different business models. Through its Open Zenium initiative, Imagine Communications is sharing source code to the Zenium microservices library to allow other companies to develop applications and features that are compatible with Zenium and other Imagine virtualized products.
Already deployed for years by big conglomerates like Amazon Web Services, Facebook, Google, Microsoft (Azure) and other online giants, the concept of serving many simultaneous sessions quickly via microservices enables these consumer-facing companies to compete on their myriad of service offerings. It simply could not be done without provisioning their massive hardware and software processing requirements to run across thousands of machines in the most efficient way. As they see it, a single compute server can be utilized by many different users (in a session) in the same day.
This same “micro service” idea is now being marketed to the Broadcast industry to help customers keep up with the increased need for IP-based video streaming. Many say the industry needs to catch up to its OTT competition, and fast.
A microservice is built on a software stack of signal processing features and then delivered via IP within a special packet bucket called either a Docker (Microsoft’s term within its Azure platform) or CoreOS (a Google initiative which stared as “the Rocket” open source project) container. Information Technology system operators use these containers to run and manage apps side-by-side in isolated containers to get better compute density.
Broadcast industry suppliers like Avid, Evertz, Grass Valley and Sony use the Docker container to deliver their technology as a microservice, while Imagine supplies whatever the customer wants, including CoreOS, Docker and a container version it developed itself that runs on any operating system.
For example, Avid’s MediaCentral platform (one of the first such virtualized platforms introduced to the industry several years ago), integrates newsroom, sports, post, and live production workflows in a tightly connected and scalable virtualized technology platform that encourages users to add additional components and apps, from third-party services to Avid’s broadcast solutions. This, the company said, ensures content is delivered quickly and efficiently to any device.
Grass Valley’s Stratus provides a set of production tools in one application while also allowing the user to better manage media workflows in order to stay competitive.
Grass Valley’s Stratus, now deployed at more than 200 production facilities around the world, is not technically a microservices platform, but it does leverage an all-software architecture to make getting productions to air easier and faster than ever before. It provides a set of production tools in one application while also allowing the user to better manage entertainment, on-air operations and news production media workflows in order to stay competitive. Since the Stratus platform’s task-driven tools operate in a variety of environments, and can be specifically tailored for each user's job function, users can increase efficiency and enable more effective collaboration throughout the entire production lifecycle.
Taking A Cue from Consumer-Facing Organizations
However, while companies like Amazon Video, Apple TV, Netflix and others are essentially delivering files to consumers’ homes using this method, the network topology that content distributors have to implement is different. In serving up video streams, distributors have to aggregate live and on-demand content that is being ingested in different file formats and data rates. This creates a more complex problem to solve.
“Wouldn't it be nice if we had a framework and we could leverage the same tools as [Amazon, Netflix, etc.,] to get similar results,” Eksten said. “We’ve built a microservices framework and a series of platforms to do that. Then we filled it full of the individual microservices that broadcasters need. What’s exclusive to what we do—because we went through the whole process of creating a new platform (based on a wide spectrum of operating systems)—is that we have maximized the utilization of the available compute and resources.
Imagine Communications’ cloud-native Zenium platform is used to customize, modify, design and deploy workflows and services with efficiency, scalability and performance.
“That’s why we’re sure customers will migrate to microservices,” he said. “We can fill all of the corners of available compute, so that we can keep costs in line and be competitive from a technology and cost perspective, which is what our customers want from us. Cost and efficiency matter.”
For its part, Evertz has introduced evEDGE, what it calls the “next generation platform for SDVN solutions.” The company said its Software Defined Compute and Routing platform, which serves up processing tasks on-demand, meets the needs of smaller facilities looking for a cost effective way to migrate to IP, while enabling larger facilities to expand their reach using a distributed environment. The evEDGE platform can operate in a standalone or core switch environment featuring redundant switch fabric technology with twice the compute, processing, and gateway density of other solutions.
The evEDGE portfolio supports bulk processing modules (FPGA accelerated compute blades) that serve as IP media gateways (supporting formats such as ASPEN, SMPTE-2022-6, NewTek NDI, RFC 4175, and AES67); multi-viewing; and video/audio conversion and processing. The evEDGE compute blade enables a virtualized environment, providing customers with access to on-demand processes that facilitate agile workflows. The company said this multi function platform allows for efficient use of hardware resources never before seen in traditional SDI installations. You pick the topology, Evertz, with over 40 global SDVN installations worldwide, says it can deliver it in software that is easily accessed via an IP infrastructure.
The evEDGE portfolio supports bulk processing modules (FPGA accelerated compute blades) that serve as IP media gateways, multi-viewing, and video/audio conversion and processing.
The Drive To Save Money
As broadcasters adapt to the changing nature of today’s media landscape, they must address new requirements of ever-changing codec standards, increased bandwidth and channel counts as well as differing video formats and the movement of generic IP data traffic across networks. A microservices architecture might be the “killer app” solution for a wide range of file-based environments.
The Walt Disney Company, a major customer of Imagine Communications and owner of the ABC Network and its eight O&O stations in the U.S., saved lots of money by moving to a software-centric content delivery platform. However, that’s not the real reason they moved to a software-defined network, according to Eksten. They did it because they can generate more content and turn up services (channels) faster, often in hours or minutes.
“At the end of the day the drive is to save money,” said Imagine Communications’ Eksten, “but also there are a lot of byproducts of that. Microservices give you much more control over your operations. You can shop around for the datacenter that best fits your needs, instead of the closest datacenter available, for example.
“Technology is not going to stop changing,” he said. “Now that we’re on COTS hardware (switches and compute), you can’t avoid it if you want to be competitive. With a microservices approach, you can define what you want your cloud or SDN network to do and how long you need it to do that. I think once broadcasters get on the same sort of parity with the enterprise side of compute, in terms of technology, microservices will be embraced industry wide.”
You might also like...
IP Security For Broadcasters: Part 3 - IPsec Explained
One of the great advantages of the internet is that it relies on open standards that promote routing of IP packets between multiple networks. But this provides many challenges when considering security. The good news is that we have solutions…
The Resolution Revolution
We can now capture video in much higher resolutions than we can transmit, distribute and display. But should we?
Microphones: Part 3 - Human Auditory System
To get the best out of a microphone it is important to understand how it differs from the human ear.
HDR Picture Fundamentals: Camera Technology
Understanding the terminology and technical theory of camera sensors & lenses is a key element of specifying systems to meet the consumer desire for High Dynamic Range.
IP Security For Broadcasters: Part 2 - The Problem To Be Solved
By assuming that IP must be made secure, we run the risk of missing a more fundamental question that is often overlooked: why is IP so insecure?