A cloud in the desert

Last month, BroadcastPro Middle East, in conjunction with Imagine Communications, hosted a roundtable on the cloud in the region. The roundtable was well attended by CTOs and senior executives from various broadcast entities in the GCC. Vijaya Cherian brings you a comprehensive roundup of the discussion, with insight into some of the chief concerns raised by regional broadcasters and a possible cloud alliance proposed at the discussion

Cloud has caused a stir in the broadcast industry. While the IT industry has taken great strides on this front, the potential of cloud still remains fairly unexplored within broadcast. Thus far, the concept of cloud has primarily been vendor-led within broadcast; then again, only a few vendors have made much progress on this front. The frontrunners are having to invest heavily and are trying to pass their initial costs on to the end user, namely the broadcaster. Therefore, prohibitive costs, the shortage of proven solutions in the market and the lack of options across the entire cloud chain have hindered the adoption of this technology in the Middle East and North Africa.

Perhaps one area where broadcasters have been comfortable attempting any sort of tinkering has been Disaster Recovery. However, the associated costs, lack of a proper licensing model and the paucity of adequate support from all of the vendors in the broadcast chain has made this challenging for end users.

We had a full house at the roundtable, with Stephen Smith, CTO of Cloud at Imagine Communications, moderating the discussion along with Anas Abu Hantash, Business Manager, Solutions Architect at Imagine Communications.

Participants included Mohamed Abuagla, CTO of Al Jazeera; Dominic Baillie, CTO of Sky News Arabia; Haitham A. Zaidan, Acting Director of Engineering at Al Rayyan Satellite Channel; Saleh Lootah, Head of TV and Radio Engineering at DMI; Afzal Lakdawala, Head of Technology Planning and Projects, DMI; Omar Alzoubi, Systems Engineering Head, DMI; Mahesh Jaishankar, VP of Datamena and Broadcast, du; Laurent Tescari, Director Product Development at E-vision; Mohamad Fares, Head of Broadcast Technology at Qatar TV; Nick Barratt, Senior Broadcast Manager at MBC Group; Paul Green, Senior Digital Technology Manager at OSN; Peter Van Dam, Acting Director of Broadcast Technology at Abu Dhabi Media; and Yves Durieux, EMEA M&E Domain Lead from HP.

The participants first decided to revisit the definition of a cloud, owing to the many different interpretations associated with it. All of the end users at the table agreed that most of them did have some sort of a virtual environment, tantamount to a private cloud, at their premises. They chose, however, to term this ‘virtualisation’ and commented that this was perhaps the first step towards moving into a complete cloud environment, but must not be mistaken for cloud itself.

Regional TV viewing trends

Stephen Smith of Imagine Communications set the stage for the discussion by first asking the participants about the demographics of their audiences. He commented that global trends show that 50% of people under 35 consume traditional linear content, while the rest have moved to on-demand or online platforms. By the same token, most people over 35 continue to consume content primarily on linear platforms. He questioned if the same was true for the Middle East as well.

Al Jazeera’s CTO Mohamed Abuagla shed some light on the demographics in the region, based on recent studies undertaken by the network. He pointed out that there was much greater diversity in the demographics than they had initially anticipated.

“We have a host of channels that target different demographics. We have various categories based on the statistics. Up to 18 or 18-24 seem to have a different viewing habit, and these are the people that are going extremely mobile. 18-35 is somewhat linear, but we have seen a gradual shift from linear to online in this age group, while those above the 45+ bracket seem to prefer appointment-booking programmes.”

Most of the others commented on the demographics related to their specific offerings. OSN’s Green said that content was also a consideration for online platform take-up. OSN noticed that during live sports broadcasts of cricket and golf, there were significant spikes in second-screen usage.

All participants agreed that viewing habits seem to be transitioning from traditional, content-based viewing to mobile and on-demand. The obvious follow-up question was how broadcasters were looking to recapture an audience that had shifted to mobile devices.

Re-strategising early on, commented Abuagla, explaining that early studies tracked that mobile would overtake linear viewing in 2015 regionally. That shift, however, happened in 2013 – almost two years early. This meant the network had to re-strategise how to offer news bulletins to people watching on mobile devices.

“I think the stickiness is probably closer to the entertainment side – sports, movies, drama series and so on – to get the smaller screen experience, but for news specifically, we had to repackage the way we do this and build platforms specific to the demographics and geographies involved. This was based on the timelines and the attention span of different age groups, hence our AJ+ fully digital channel,” he added.

Transitioning from a linear to an IP model – the implications

Smith then headed into the core topic of the day by commenting on how the marketplace is changing; new delivery mechanisms are becoming gradually available with the transition from a traditional linear overview to an IP delivery model. Of course, this gradual transition has opened the doors to creating a lot more channels targeted at specific events or maybe, just split in terms of targeting a regional/micro-regional demographic.

“One of the ways to do this is to move out of a physical infrastructure into an elastic model, which means you are not just being provided with basic infrastructure but are partnering with someone who provides you with compute, connectivity and storage on-demand. You’re launching these services as pure software solutions as and when they are needed,” commented Smith.

While Baillie agreed with him, the CTO pointed out that one of the basic challenges lay in the fact that a large number of geographies that local broadcasters broadcast to do not have the required connectivity. He pointed out that satellite was still the most effective way to reach many people.

“Satellite is a cost-effective way to get your channels across to these people who do not have proper connectivity. I don’t know when these smaller towns in Saudi Arabia and so on will have enough capacity to access this.”

Mohamad Fares of QTV seconded this by saying that a large portion of Qatar’s homes have a 1MB connection and no fibre connectivity, making satellite the most viable method for broadcasters to reach viewers. It was agreed that the problem was not confined to the GCC area but spanned the MENA, including the Levant. Most agreed that connectivity was more a geographic issue than a demographic one.

With this perspective in mind, Smith asked if originators would consider moving to a more software-oriented solution.

“Regardless of whether we are trying to create multiple IP channels for targeting micro demographics or an appointment-based linear delivery over satellite and FTA – if we can use the same technology to target all platforms, and instead of having that technology exist in your own data centre, it was hosted in a public or private cloud, with everything running in software, and your channels are delivered to an earth station or transmitter as a transport stream over IP, would that be something you would consider?”

Peter Van Dam from Abu Dhabi Media pointed out that legal issues would be a huge concern for the state broadcaster.

“To move our content to another cloud somewhere that is not government-endorsed would be a huge concern, so legally, this would not be possible,” he stated.

Saleh Lootah, Head of TV and Radio at Dubai Media Inc. commented that if it stayed within the country’s borders, that might work.

Public cloud service – Does it exist in the Middle East?

This brought the team to a more primary question: Where is the nearest instance of a public cloud service provider in the region?

Mahesh Jaishankar, VP of du Datamena, agreed there was nothing like an Amazon or Azure in the region, although Datamena came the closest to offering a potential virtual environment.

The group, however, argued that the available services were way too expensive in terms of IP connectivity and latency.

“The lack of a public cloud service such as the one provided by Microsoft and Amazon means we do not have the infrastructure or the kind of opportunities associated with those services. Someone in Qatar offered me access to Amazon services through a VPN to Europe. That’s sadly the kind of opportunities we have here,” Baillie commented.

Virtualisation – the first step to a mature cloud environment

Some end users in the room, claimed to have taken the first step to the cloud with a virtual setup.

Mohamed Fares, Head of Broadcast Technology at Qatar Television, pointed out that the broadcaster’s “content is hosted and broadcast from both a private and a public cloud for different applications”.

“QTV virtualises a significant amount of its application on an infrastructure as a service platform. One of the successful implementations is the virtualisation of Imagine’s Broadcast Master platform on a completely redundant, highly available cluster. These measures of operating on a private cloud have given us great benefits in terms of an elevated viewer experience, decrease in total cost of ownership and decrease in operational complexity,” he elaborated.

Participants questioned if the QTV implementation was indeed a cloud implementation in the full and mature sense of the term.

“Not yet,” Fares conceded. “In some cases, it is a virtual server, but we do also have some physical servers. We look forward to seeing the technology evolve and allow us a higher level of integration in terms of operating on the cloud.”

Baillie pointed out that most broadcasters do already have such instances of deployments within their premises. However, he added that the majority of broadcast vendors are still unable to provide the support required for end users to virtualise their servers.

Haitham Zaidan, who had just deployed a virtualised environment with Imagine Communications, agreed that this was the case.

“I come from an IT background so I have been wanting to virtualise my environment. When I approached Imagine, they said they didn’t support it but I went ahead and did it myself although now, they have deployed a proof of concept at our facility,” he commented, clarifying that it is still only a virtual environment and not cloud in its purest form.

Lakdawala of DMI commented at this stage that every broadcaster is using “some form of private cloud within their premises”.

“The challenge is having a public cloud in the region.”

Laurent Tescari from E-vision added that every telco is aware that IPTV is becoming increasingly important, but they have to be convinced there is a business case.

“There is an opportunity for some MENA telcos to migrate to the cloud in the next couple of years. All the telcos are not comfortable in continuously investing in an infrastructure that is not necessarily doing well or is profitable,” he commented.

At this point, Abu Hantash asked if broadcasters saw themselves moving from their traditional broadcast chain to a standard IT platform rather than having a proprietary platform from each vendor.

Redefining the cloud

Al Jazeera’s Abuagla commented that the above is not a cloud offering.

“That question defines a software-isation of broadcast technologies. What I am asking for is mature software solutions that run on agnostic systems. Give me an OS and the specs, and I should be able to run anything on it. We are looking for that day in broadcast, I don’t know if that day is here yet, but it is the first step. Then you have to go into the virtualisation of that entire environment, which is the next step, and then you have to go to the cloud-isation of that whole environment again. I don’t think we are there, because there is no ecosystem that does this in a seamless fashion at present.”

He added that cloud for the IT market became more widespread and mainstream after they “figured out all of the pertinent parts of the chain”.

“They had sorted out all the steps to the cloud. They ensured that software-defined networking, software-defined data centres and so on were sorted out before, so they were able to give you a full ecosystem to work with. While I appreciate that some are trying to push the industry in that direction, I don’t think the vendors in general are ready.”

Baillie noted that delivery of content through traditional methods and new media is being handled differently.

“I guarantee you there is enough cloud infrastructure ready to support web delivery of content, so all that is going to happen is traditional vendors need to disappear from the infrastructure and completely go new media. Replacing SDI with IP is a broadcast way of keeping the industry moving forward. It is not necessarily the right way to do it, in my opinion.”

Business models – Licensing or utility models better in a cloud?

Smith pointed out that people often get buried in the details of making the technology transition and lose sight of the business goal.

“This comes back to the question of how do you manage the transition from yesterday’s infrastructure to what you think the infrastructure of tomorrow will be, which most believe is simply migrating to IP. IP is a necessary step in achieving software-defined workflow and cloud enablement, but it’s not the goal.

“It’s about breaking the model of delivering traditional channels, breaking traditional workflows and doing things differently. It is not about old media and new media; it’s about delivery platforms that can target all audiences consistently.

“People get fixed on changing the wire while maintaining yesterday’s workflows, instead of focusing on how to be relevant and delivering their content everywhere.”

Baillie argued that it could be done but queried what the cost of such an exercise would be.

“I would argue that most of us in this room do not have a huge amount of money to play with. Until you have embraced it like the US, where we are only paying for usage, it’s not going to work. I need only one ticker licence across 20 channels. I could spin up the service when I need it, use it and spin it down when I don’t need it. I want to bring out an extra channel because I am broadcasting cricket and I have six extra angles. That true micro service architecture is what we require for a cloud environment.”

Green cited another example.

“We did a project a couple of years ago looking at the cost of DR and the options of outsourcing to another site or utilising a cloud-based solution. We expected to see the price of the cloud solution to be negligible until we spin it up when required, but we found it was costlier than outsourcing to a dedicated ready system. Maybe when we get a public cloud with more regional players in and established vendors supporting software only, it will become cost-effective.”

He commented that the cost was primarily a combination of different factors. In addition, achieving it in the cloud was challenging because many vendors do not have licensing models to support it.

“We spoke about if we own licences, these should be transferable between our main and DR site; we shouldn’t be paying double for something we would only be using exclusively. Some vendors may give you 50% discount for example, but that, to me, is not acceptable in a software-only model. When we also looked at hosting services on the cloud within the DR scenario, most of our current vendors did not offer a software-only solution that we could potentially run in a virtualised environment. So then you deal with the challenges of having a different set of vendors in your DR solution and the associated complexities of training operators using multiple systems, one of which is never hopefully going to be used.”

He added that if people were provided a choice between a ready-built site and a cloud environment, most people would opt for the former because it runs perfectly.

“There has to be a reason to move to the cloud, and that should be that the cost of moving into the cloud is negligible until you spin it up, whereas with this one, I have to pay capex and build it up.”

Baillie pointed out at this juncture that it would be more sensible to build a new system in the cloud, and then the DR would just be additional instances. Green agreed, but pointed out that at as a large pay TV operator, for OSN, this could be risky.

“We have a large customer base and trusting a new technology in the region where we face so many challenges is extremely risky. Doing it in a DR scenario is less risky, as you can build confidence in the solution before moving into your primary, but even that is a challenge because it is not currently cost-effective,” he said.

A software-based model – The way forward

Smith pointed out that if you consider only a portion of the environment instead of looking at the bigger picture, it will always appear to be more costly.

“I would argue that virtualisation is a transient technology that allows existing and legacy applications to function in a cloud environment. We really want to avoid the concept of a virtualised server running on a specific host. Instead, we want autonomous micro services that have no notion of a machine at all, no notion of local storage or that you own a fixed amount of compute. It should be a simple model of consuming resources – resources that are either compute, connectivity or storage and are not dedicated to a specific machine.

“When we look at DR capabilities, we should not be thinking of a DR system or site, as this implies we are writing multiple cheques without gaining additional functionality. You write a cheque for your primary chain of origination, and a similar-sized check for your backup chain of origination, and a third check for DR, and you’re hoping you will only ever have to use the primary chain.

“In this model, two-thirds of your infrastructure remains unused although you pay to keep it running and for staff to manage it. Instead of looking at redundancy, we need to look at resiliency. Instead of a primary site and a DR site, we need to assume that all resources are equal and that it’s simply a big pool of resources. Now there is no notion of a primary video server and a backup video server, a primary graphics device and a backup device. Business functions are protected based on configuration and priority.

“We should scale for business as usual, with some additional capacity in the environment to allow for degraded performance in some portion of the infrastructure. In the event of an outage, you can reduce capacity for a less important business function to augment another. That is what cloud is all about — dynamic scaling of business functions to align with the current business needs.

“If you have a DR facility and you own it and run it yourself, you should not be thinking of it as a ‘just-in-case’ infrastructure, but as an extension of business as usual. You have paid for it, you are depreciating it and you should be utilising those resources every single day. When in a scenario of reduced capacity, you may have to sacrifice some business functions, and you should know what those business functions are.”

Green countered that by saying that no broadcaster in the room had a green field site.

“If we did, we may seriously look at such a solution. It would be a good way forward, because you have a single main and DR system, and can train everyone once. But we have an established system, where we replace aspects every few years as technologies advance, but to move everything into that environment and have a single setup is a challenge.”

Baillie seconded that by commenting on the risks involved. “You have to be able to take some sort of controlled or phased approach. Ideally, you should have a whole system in the cloud and slowly transition your content to it. In reality, however, having worked with at least three broadcasters in this region, the ability to do that is hard to come by.”

Yves of HP commented that cloud is indeed a journey.

“It is good you are looking at where you want to go and thinking small steps. DR, however, is a good way to start.”

Baillie interjected at this point that no one is disagreeing.

“Taking a small step or moving parts of a system to the cloud is not the problem. The issue is money. It is prohibitive. I have to invest in a private cloud. This is only cost-effective if it is public cloud. I don’t have a POP here, and therefore I have connectivity and latency issues getting out of region.”

Green agreed cost was an issue even for taking a first step towards a cloud solution.

“In the US, people transcode on the cloud because it is cheap and flexible, but it works because there is cheap connectivity. We looked at the option of sending our high-res files to the cloud to be transcoded. The problem was I have to pay du too much to get it to them,” he commented.

Call to Action – Forming a cloud alliance in MENA

Abuagla seconded this.

“It is amazingly costly to network any two points in the Middle East or the region in general. It costs a crazy $15,000 to network one point to another, versus the $1,000 which will get you that T1/E1 line in the US. Co-location is also hugely expensive here. This is an evolution that will take time. We need to work together. I don’t think broadcasters will be willing to jump in for several reasons including legal issues, risk, cost, connectivity and the fact that our business does not have the tolerance to test new technologies live. I have the tolerance to play with them, but give me a playground to familiarise with.”

“I don’t see the industry even interested to build a free community cloud for broadcasters to play. On the IT side again, there is a cloud alliance. We need to have one for broadcasters so we can all operate under one umbrella.”

Abuagla’s call for an alliance to create a “playground” was welcomed instantly by all the broadcasters. “I think if there is such an alliance that basically allows all the really big vendors to come together and show the broadcasters an ecosystem that works end to end, it will be more effective. Show me how a complete ecosystem works. Secure it. Address legal issues. Then I will play, but I am not going to pay for the right to play; however, I’ll pay for what I consume.”

Du Datamena’s Mahesh Jaishankar offered to host the playground for everyone if the vendor community could be engaged to come together to support the exercise and provide the required connectivity.

Abu Hantash commented that Imagine Communications was already in talks with HP, Datamena and other players to create a consortium. Everyone agreed that there were several virtual environments in the region but no proper cloud case studies.

“I haven’t heard of a complete solution around a true private cloud that actually mirrors the characteristics of a cloud as a service model. I need to be able to use it, charge it back or show it back. I need to be able to know I can consume it and meter that consumption. I should never run out of it. I can use it as much as I can, and I can push it more and it expands, and I can push it even more and it still expands. I don’t think anyone has that here,” commented Abuagla.

DMI’s Alzoubi suggested a hybrid cloud environment, which gives the end user the flexibility to operate parts of his environment in a cloud.

Cloud as a service

Smith defended virtualised solutions, explaining their purpose in the transition to cloud.

“For me, cloud is technology in an as-a-service business model.”

For the most part, he conceded that when talking of the cloud, the reference was primarily to “virtualised systems running on a private infrastructure”.

Al Rayyan TV, which recently undertook a private cloud at its office in Doha, commented that the decision to move into a virtual environment was planned two years ago.

“We virtualised our MAM, traffic and scheduling system, workflow management and data bases. Next year, we will move editing, graphics, automation clients and all users clients to a virtualised environment. We saved a lot in the last two years in terms of support time, rack space, even power consumption. For example, we had more than 10 servers to host Imagine products but now, we are using only two physical servers to host the same number of virtual servers,” stated Zaidan.

Lootah from DMI, however, questioned if loyalty to one brand was a smart move.

“What if the company that has deployed your entire system goes out of business? You don’t have a support contract. I would be concerned about running my entire virtualised environment with just one vendor.”

Abu Hantash responded that even in such an instance, “the end user would lose only the software part rather than losing one’s investments for both hardware and software if they move the standard broadcast chain (media, automation, video server, GFX, routing etc) into a 100% software model running on commercial off-the-shelf IT platforms”.

“This would enable rapid deployment, simplified access, monitoring and control, configuration over installation, with no equipment on the customer’s premises or investment required compared to what we do now.”

Abu Hantash also defended the Al Rayyan installation, saying that the virtualisation would have happened at the broadcaster’s facility anyway.

“They already had an HP Blade. We created a couple of virtual machines and we installed our VersioIP and got the IP stream output. I was support engineer, and it took me two weeks to commission the traditional system with the huge rack spaces, wiring, etc. This is something that will happen sooner or later. The best practice for us understanding the IP/Cloud solutions is now to deploy proof-of-concepts and see how broadcasters see the benefits. Based on their individual feedback, we can help build something together for the end user. We are also working with HP and du to see whether we can build POCs in each facility.”

The discussion then moved to how much downtime was permitted within a broadcast environment.

“Cloud service providers will offer a fixed SLA, which may be okay for your email server, but we need to see if it is suitable for a broadcast environment where, for example, during a Game of Thrones finale, downtime is unacceptable,” commented Green.

Abuagla took the discussion one step further by stating that “the cloud perspective should take away the whole need to DR or BC”.

“The cloud should give you an SLA and an availability standard, and then I shouldn’t have to worry about it anymore.”

Everyone questioned what kind of compromises were acceptable during a downtime. Lootah pointed out that confidence in the cloud was so low at the moment that he would accept a virtualised channel in the box only “if my primary is gone with all of its redundancy, if my DR in Studio City is gone and my MediaDeck in Samacom is gone”.

‘Then, perhaps, I will accept a channel running out of a box with limited capabilities, because something is better than nothing. But I will first exercise all the other options available to me. It is just cutting all three steps in the middle and moving.”

Abuagla suggested working from the outside in.

“If the vendors can prove that the cloud environment is safe and works, you may move it to the third tier, then the second, then the first. That would be the evolution you are talking about.

The Al Jazeera business model

Abuagla explained Al Jazeeera’s business model at this point, which most agreed was hugely impressive.

“We have a broadcast centre in London and DC; we have operations working out of New York, Turkey, Balkans and so on. By design, we are obviously going through a massive upgrade of a lot of all our infrastructures. We have built it in a way that our DR for Doha is London and the DR for London is DC. The DR for DC is KL and the one for KL is back to Doha. We pretty much have done it, but I own the network and I had to pay to connect everything absolutely fully, with a lot of money.

“That’s why the big deal we did with Ooredoo facilitates our cloud future, but we had to write out several cheques for each setup in each location. You’re building that promise within your environment, so I could call it if I connect it, and I guess orchestrate it. I can build it as a big private cloud environment that is truly serviced based on all of this stuff, but this is a luxury others don’t have. I have this because I have to be geographically dispersed and I have business models to which I have to respond to out of the US, London, etc. I can build a DR solution that is basically an actual production solution for someone else.

He added it would have been ideal to have seven primary sites and one DR site, but “there was nothing available”.

“If I had the ability to build my core out of Doha, I would have done that and, basically, built cloud instances for London and cloud instances out of the US. If the industry was capable of giving me those options, I would never have had to write that cheque multiple times.

“We are building a mini, private cloud; we are building our own infrastructure and systems, but we are also building our own search capabilities so anybody anywhere should be able to access it without any issues. I am, however, willing to divorce from that if I have a great cloud solution.”

Smith commented that this is the path Imagine is headed down.

“If you just look at the intent of cloud, instead of scaling up or scaling down, we are looking at n number of licences. You may buy 10 licenses and you may decide to scale up to 100 channels and have to buy 90 more, and then what happens when you reduce it to 90, and what are you going to do with those 90 licences? So we have moved to a utility model. So when you start the channel, the meter is running, and when you shut it down, the meter stops, and this is being fine-tuned right to the last level.”

The broadcasters called for a shopping cart experience where they can pick and choose the solutions they want without having to worry about the infrastructure that drives them. Abuagla pointed out that the IT industry has already achieved this progress by sorting its infrastructure service, moving into databases, applications, authentication, analytics and so on, and now it has gone one step further by looking at platform packaging as a service type of model.

“Essentially, you have a shopping cart, in which you put all the items you want … they are pre-integrated and have all the API figured out between them. Everything is software-defined and only when that has been done have they moved to the next step,” he commented, adding that everyone is cognisant that the sensitivities on IT are different from broadcast.

He went on to question whether there were enough vendors buying into this model and actually transitioning into virtualising their software so it can run on this model.

Smith countered the arguments, stating that end users should not be picking and choosing systems. Instead, they should look at what they want delivered, as in 2 MPEG transport, delivery of three languages, graphics plus capture and so on.

Should cloud be vendor-led or user-led?

Nick Barratt of MBC Group pointed out that the difficulty does not lie in specifying the functionality, “which is what you want to do, and you have this in any sort of hardware replacement”.

“It is the back-end functionality that hits us, and it’s the same change you have to manage when you are swapping out one vendor’s solution for another.”

A conversation raged on whether there were too many specifications on specific vendors, solutions and so on. Smith argued that “all these systems exist because you want to bring that complexity to the table” and advocated a strong software approach.

“You have physically dictated the chain you have to n number of times to replicate it. In the software world, you couldn’t care. You would care about the functionality, and your specifications would go like this – here is the template for my channel and for resiliency. I always want it to be two, and I don’t care about where they are originating from physically, but I need to ensure they are not put on the same physical machine, or the same physical network or data centre. Then you may be paying only for one package. You probably are paying for what is currently being packaged and received.

“If we abandon this notion of I have all these disparate systems that are connected by unidirectional coax, and instead go with a software model, then there is really no automation anymore.”

Abuagla agreed, stating he does not want discussions to revolve around vendors and products.

“I want them to be around features and solutions. I don’t want to say I want this vendor and this product and plug it into this and that. I want to be able to say I want these ingest capabilities; I need to have archiving and editing capabilities. I need to playout on these platforms, and I need analytics; and here is my SLA.”

“I don’t care what technology you are playing in the background or what storage is being provided to me. I don’t care how the chassis looks and what hard discs are installed in this thing. I just need to support my business operations, while cloud vendors provide me the tools to do so.

“I want to ensure I want storage whenever I want it. I need the following services regardless of who is providing them, fully integrated, and they will give me this thing, and I want to pay for what I use when I use it.”

Anas interjected that the one-stop-shop Abuagla requires “doesn’t currently exist”.

“Maybe! But when that happens, we have the cloud,” argued Abuagla.

Conclusion

Smith concluded by commenting that the transition from traditional broadcast to a more personalised and immersive experience is definitely happening. He added that the current step in that evolution is to change the wire from coax to Ethernet, although this also allows you to get to a state where you can begin to virtualise, which is the first step in getting to cloud enablement.

“Once you are there, you can then stand up for new channels, looking at changing business models, changing revenues in the market, etc.”

Barratt was quick to add here that he was sceptical of being vendor-led.

“The propositions that are coming to the table are single vendor-led and that limits choice,” he cautioned.

Lootah called for greater collaboration between telcos and the vendors, while Baillie added that for content consumption, we need to look less at traditional broadcast suppliers and look at the likes of services providers within IT.

“Everything we have discussed is largely on the consumption side, and I don’t see a future for broadcast manufacturing companies in that side. If you really want to make the transition, you have to sit there and rethink what we are trying to achieve and what’s the best way to get the decision-makers to take a positive call. The way we have done things thus far is not necessarily the best way. It has evolved through need and technical advancement.”

The roundtable concluded with some key points: a virtual private cloud environment is not the same as a public cloud experience; local broadcasters require a playing field where they can familiarise themselves with the different cloud features; and the vendors and service providers at the table (Imagine Communications, HP and du Datamena) would aim to work out a strategy that would include more players in the future.