Day Two Keynote Presentation by Alessandro Talotta, Chairman and Chief Executive Officer, Telecom Italia Sparkle
Unlocking European Economies with the Mediterranean Cloud
The second plenary session of the NetEvents EMEA Press & Analyst Summit was opened by a keynote speech from Alessandro Talotta, who talked about TI Sparkle’s networks and datacentres in Sicily, Greece and Turkey, and how they are changing the Mediterranean area by creating new markets, empowering enterprises, speeding cloud services, and boosting the local economy.
He started by talking about about the historical importance of the Mediterranean basin – where communications roads meet between Asia & Europe. It’s the same with network cables, he said. Sicily is the jumping off point for cables to the Middle East and Africa.
He said TI Sparkle’s main network hubs are in Sicily, London and Marseille, with managed bandwidth growing 44% year-on-year. TI Sparkle’s biggest datacentres are in Palermo, Athens, and Istanbul.
He talked about the service portfolio, which includes PaaS, IaaS, SaaS and datacentre services such as DR, hosting and so on. Taletto said he sees TIS as an open place to peer networks.
He was then interviewed on stage by Ovum lead analyst Camille Mendler.
Debate Session IV — Telco “Hot Trends”
Introduced and Chaired by: Ian Keene, Vice President, Gartner/em>
Panellists: Gary Bolton, Vice President Global Marketing, Adtran; Kamal Okba, Director, Maroc Telecom Group; Kevin Vachon, Chief Operating Officer, MEF; Phil Tilley, Senior Product Marketing Director NFV and CloudBand Solutions, Nokia
Ian Keene outlined key trends, including LTE-U, 5G, cloud radio area networks, smaller cells blurring fixed and radio access, digital business transformation, and SD-WAN (SDN & NFV).
He saw new market developments: Internet of Things (IoT) will open new competition in the mobile service market. We will see a more application centric market, he said, with more digital business transformation, and Internet access via mobile.
How do operators / service providers respond? Lifecycle service orchestration (LSO) is the answer – it sits on top of SDN/NFV and the MEF will tell us what it is.
Kevin Vachon explained what LSO is – it’s about services driven by need to provide high speed bandwidth on demand. As volumes scale as need for agility and reduced opex grow, the need for LSO grows.
Phil Tilley said we will see more customer self-service, more IoT, and competitive pressure all driving an increased need for automation to reduce operational costs. LSO is about how you orchestrate end device connectivity into the enterprise, so allowing the end user to select services from catalogues, he said.
Gary Bolton said we see convergence of all sorts of operators into service providers with a ubiquitous need for connectivity. The self-directed consumer is one who can push a button and get what they want. Also we see customers having to change processes to become more agile so as to launch new services very quickly. If you don’t so that, Amazon and Google will eat your lunch. Everything becomes commoditised so fast that innovation is key. An example is Huntsville, Alabama, which laid dark fibre to improve connectivity and attract Google. The key is to innovate on services not just provide fat pipes.
Tilley said Google is in very strong bargaining position when negotiating access deals.
Bolton said it’s not just about pipes but changing policies to make connectivity cheap and to open up networks so franchise agreements are standardised.
Kamal Okba asked if there were not issues when moving to customer self-service. He said Google should apply for a telco licence and respect those obligations but that telcos generally need to be faster and more agile. OSS needs improved architecture to be more secure. Today operators need solutions to help them reduce opex and capacity to expand quickly. Much needs to be done – in Africa we wait until initial problems are resolved – we have existing investments and can’t move quickly. The priority is OSS – new standards will make the lives of network engineers easier.
Tilley said managed transformations are difficult – but new engineers are into building new apps on Linux so coding is now a part of life. They do things differently.
Bolton agreed it was a difficult shift because telcos have to obey a set of rules to provide services to everyone not just cherry-pick the most profitable.
Vachon said the industry has hit a wall. Engineers need to build new services quickly – people can’t wait months for a phone to be activated. Customers want to self-activate and get services turned on quickly. Everything takes longer than you think.
Analyst Dean Bubley from the floor asked: will telcos look like AWS in future?
Tilly said services will be able to be switched on quickly via automation so setup costs are low. Short term VPNs will open up, also enabling hybrid cloud.
Bolton said that convergence of networks mobile, local loop, etc. will all deliver data such as TV over any device. It will be cheaper more seamless, offer more choices, and be open access.
A question from the floor asked if the lack of difference in terms of service delivery and SLAs between business and residential services will continue.
Bolton said this seamless single network will allow you to choose what’s important to you. You just pay for what you need.
Vachon said the MEF Third Network concept we released 18 months ago will deliver that single connection no matter where you are. The need is there.
Tilley said that SDN will help operators deliver better service adding services and features, and security policies as required.
Debate Session V — Enterprise Dilemma: Outsourced Managed Services vs In-House
Introduced and Chaired by: Bernt Ostergaard, Analyst and Service Director, Quocirca
Panellists: Kathy Schneider, Senior Vice President Product and Marketing, EMEA, Level 3; Chris Lewis, Telecoms Industry Analyst, Lewis Insight; Kamal Okba, Director, Maroc Telecom Group; Paul Davies, Director of EMEA for Cloud Services, Verizon
Bernt Ostergaard said he would adopt the role of an enterprise CIO and ask how to get the best service. How well does corporate IT shine? Most (85%) need to raise their game, improve competency, manage multi-sourcing better. Defining which tasks need to be outsourced – frameworks can help, he said, to manage integration issues, service desks and security. The CFO will ask me (as the CIO) to justify in- or outsourcing decisions.
Kamal Okba said that in countries where Maroc Telecom operates, he finds mixed expertise. In N Africa, outsourcing usually happens for installing broadband or fixed lines, or problem solving. From Maroc Telecom’s perspective, the enterprise needs first to decide which activities to undertake in the first place – if an activity has no impact on performance and/or is of low strategic importance, it should be eliminated or outsourced.
Kathy Schneider said we focus on network and managed services including content delivery, all within context of Level 3’s portfolio. She said we want to add value to the foundation of the business: the network. The work we’ve done with the MEF is very exciting because people want more services, she said. Automation helps to enhance network services from delivery to billing.
Paul Davies said large global enterprises are transforming their networks because they have become a bottleneck to success. We can help with that by offering eg managed security services. We bid for work that systems integrators might normally do because enterprises see service providers (such as Level 3) as working hard to deliver, he said. We can now provide managed cloud services following the Terremark acquisition, he said.
Chris Lewis said there’s a lot of conflicting dynamics – each component thinks it’s the most important, but it’s the combination – not just the network etc that is important. The CIO wants to support the business not just run IT – they don’t want to be tied down by old telco practices and they need managed services to connect remote workers and branch offices. It’s slower than may would like – is this the result of conflict between network and application providers? The orchestration management issue is key, he said. Service providers are like kids in a sweetshop fighting over the sweets where they should be more open, more customer-centric, making it easier to use their services. We have a love-hate relationship with operators from both the consumer and enterprise points of view – they need to deliver better value.
Okba said telecoms providers are not the devil – we take our responsibility seriously.
Schneider said customers are saying that they know how they want the network to behave.
Ostergaard said enterprises are saying their different connections don’t play together and that provisioning times are too long.
Schneider said that as LSO comes along, traffic will converge inside the network.
Lewis said there’s lots of bandwidth out there, it’s just not being used very effectively. Customers may not know what they want but they know what they don’t want. Why do service providers outsource customer support centres when customer service is supposed to be strategic? Most of what SPs do is horizontal eg Vodafone acquiring Cobra and getting a handle on automotive services.
Ostergaard asked if the MEF was the right place for discussions between carriers and customers.
Schneider said yes, we have the highest number of MEF certified professionals. The MEF helps us get different points of view and perspectives, helps you move out of internally focused company culture. There’s still lots of functionality to add.
Special Guest Speaker Presentation by Jeremiah Caron, Vice President – Analysis, Current Analysis
The Internet of Things: What’s Really Happening Now?
Debate Session VI — Goodbye Earth! What’s the future for hardware vendors in a virtual universe?
Introduced and Chaired by: Pim Bilderbeek, Analyst, The METIS Files
Debate Session VII — “Stress testing” Cloud Applications and Infrastructure
Introduced and Chaired by: Steve Broadhead, Founder and Director, Broadband Testing
Panellists: Jan Guldentops, Director, BA Test Labs; Roark Pollock, VP of Solutions Marketing, Ixia; Paul Davies, Director of EMEA for Cloud Services, Verizon
Steve Broadhead asked: What’s changed over the last 20 years – how do we test and should we test? He explained how he tests. He asks: does it work, how long before it breaks, how does it compare, and how does it perform – and what about scalability? Broadhead showed examples of the kinds of tests his company ran over the last 20 years.
Roark Pollock agreed that the world of the test equipment vendor has changed. Software is more important in a world where products iterate in weeks. We now deliver testing as a service, on demand.
Jan Guldentops said we try to automate just as with any industry but there’s always human error to contend with. So automation is needed for the stupid stuff.
Paul Davies said there’s an expectation that our services will be rock solid so there is a need for testing. We are a big consumer of Ixia products.
Pollock said apps should be tested as soon as possible and maintain visibility all the way into real world implementation as this helps improve the understanding of real world environments.
Guldentops said DevOps means that untested apps don’t make it – and the scale has changed so if things go wrong they go badly wrong. So you need a complete testing system. But the trouble is that few people actually know what they’re doing. Like with Docker, people make it up as they go along.
Pollock said almost all our products now are available in software so they can be virtualised and run anywhere.
Davies said that because testing is software, from an enterprise perspective, the distributed network and infrastructure means there’s only so much they can test – they load-test in different geographies but they need the visibility to tune equipment.
Pollock said that customers now are part of the test process – they get beta products.
Guldentops said testing companies can be very small now – we have a great future!
Pollock said that there’s a lot to learn, a steep learning curve.