By Ian Redgewell, Head of Media Management, Globecast
We’ve all been in danger of being deafened by the noise surrounding the move to IP over recent years as the fundamental protocol underpinning our sector, alongside the associated shift to virtualized technology and the cloud. Specifically for us, cloud-based playout, of course. But what are the challenges for a global media services provider with this move, what are the real-world benefits and where are we at with IP overall?
We have spent a lot of time looking at cloud playout in great detail. For us, there are three clear strands to this: technical, operational and commercial (how we position and sell these services to benefit our customers). Taking the technical first, we are quite far advanced now along this path, as visitors to our booth at CABSAT will be able to see via a comprehensive demonstration. We have specified and are thoroughly testing all the technology currently required, including of course a cloud-based playout engine. But this is an ongoing process, not a finite one and will continue in the background as our services evolve. One of our key customers has allowed us to use its content so we can show the whole workflow using cloud storage and CDN delivery to mobile devices. All the technology being used is virtualized, be that on-premise or in the public cloud, and running on COTS. We are also now in the process of building a private cloud.
We have been demonstrating how this works to customers, showing how we can — almost instantly – spin up multiple channels including live content, and then they can see those services on their mobile device. It is impressively fast and that’s undoubtedly a gamechanger. Some customers want to see the inner workings; others don’t. There will also be a customer app for service monitoring as we move forwards.
As I say, of course we are being very thorough and leaving no stone unturned. As you would expect with a shift of this scale, there are still some things to perfect – including with live content – but the benefits are clear and demonstrable.
The technical aspects are no longer where the challenges lie. With public cloud services now benefitting from both GPUs (graphics engines) as well as CPUs, one of the key obstacles (handling graphics) has been removed. But for a service provider with our level of installed technology at our four key processing sites (Los Angeles, London, Singapore and Paris), it’s very far from being the case of simply flicking a switch. We have a huge amount of installed infrastructure that has been built up over the years and works perfectly. We are not here to impose new models on our customers. We are very carefully working through these operational and commercial strands so that we offer the most appropriate levels of service using appropriate technologies. That might sound a little trite, but this is fundamental to both our business and of course theirs.
It is absolutely true that using virtualized technology, both on-premise and in the cloud, allows us to spin up services in a matter of minutes. Indeed, we can do this faster than we can get access to the content! With the cloud, in general terms, if we need the systems to do a little bit more based on new customer requirements, then we can simply “turn up the gas”. That is new for the industry. We pay for what we use and then we can dial it down again.
That dovetails with the second key point: because this allows a pure OPEX model, we will now be able to service customer requirements to launch a service for, say, six months while they test a new market and then take it down again just as quickly, without financial penalty for anyone. Additionally, now that we’re accessing huge amounts of content in the cloud rather than it being delivered to us, tying up cloud playout with that cloud content access creates a whole new level of efficiency, both cost- and technology-wise.
As an industry, we are still at the stage of exploring additional benefits of using the cloud. One thing that really interests us is the flexibility you get with handling a large number of live feeds: in the cloud, as long as you have the IP address of a live feed, then off you go. No input/output restrictions and so on.
We are also exploring the possibilities that cloud playout offers for designing economic disaster recovery/business continuity solutions for our customers. Previously, one had to replicate the entire primary playout infrastructure and keep it running 24 x 7, and this involved high CAPEX. With the advent of cloud playout, and the ability to spin up and spin down channels quickly, along with the ability to adopt a fully OPEX-based model, designing truly robust disaster recovery solutions will become an easier reality.
In my view, the vendors are there with IP products. Yes, there’s still IP standardization work happening in the background, which is vital, but that’s following an entirely natural development curve. We of course have an internal multicast IP network that services growing requirements, not least the OTT market – or huge multitude of markets – and the associated ever-expanding device support requirements.
With IP-based content processing and workflows taking center stage, customers today would like to transparently track the status of their content as it goes through the various workflow steps in a service provider’s facility. To fulfil this need, service providers are building media management portals that are tightly integrated with their back-end systems and can be used by customers to take corrective action if there is missing material, or in case there’s a quality issue with the delivered content. These portals are also able to track the status of playout and VOD delivery to affiliates as per schedule.
Again, the challenge for us – for any service provider – is to move gracefully through this implementation and to operate in what will be a hybrid technological world for years to come. Cliché I know, but central to our success is to carefully architect a specific service – or set of services – that is/are relevant for each client, supported by tiered SLAs for different levels of customer needs all wrapped into a self-service portal/app with premium UX (this is another product on the verge of production that we’ll discuss in a future blog). Yes, being as thorough as we are takes time, but we’d far rather spend that time fruitfully now than have to undo slipshod work down the line. Brave new worlds require bravery but alongside careful consideration and lots and lots of testing.
Click here to book a meeting/demo at CABSAT.