Interxion - Moving Beyond Space, Power & Cross Connects with Value-Added Cloud Connectivity Services
It was great to announce our new partnership with Interxion (NYSE: INXN) last week – one of the top Pan-European providers of cloud and carrier-neutral colocation data center services, not only because they are one of Europe’s leading colocation providers, but also because they recognize the huge opportunity to offer value-added cloud connectivity services to their customers beyond traditional space, power and cross connects.
In deploying their new Cloud Connect platform, Interxion is now able to offer to their enterprise and service provider customers, a single high performance, reliable and secure connection to multiple cloud service providers, such as AWS and Azure. By providing private virtual connections between organizations and cloud service providers, Interxion’s platform enables businesses to bypass the public Internet and achieve service level guarantees for latency, throughput and security – key requirements in supporting businesses’ migration to the cloud. Additionally, Interxion’s self-service portal allows their customers to set up cloud connections and adjust bandwidth as needed!
Key to Interxion’s selection of BTI’s Cloud Networking Solution for this exciting new service offering included:
- Density, scale and performance advantages of BTI’s converged networking platforms, with a clear migration to 100G;
- BTI’s SDN-enabled software control layer for automation and programmability with open APIs to integrate seamlessly with Interxion’s self-service portal;
- End to end SLA monitoring capabilities; and
- A comprehensive and customized support package
"BTI has provided us with a 'one-stop' approach combining: a high-density SDN-enabled converged packet optical platform; a software control layer with open APIs to enable customer self-service portals; and a customized technical and support plan. BTI's Cloud Networking Solution also offers a clear migration path to high-layer intelligence and 100G capacity." said Interxion's VP, Product Management and Channel Development Mario Galvez.
The announcement underscores the opportunities for forward-looking colocation providers who are seeking innovative ways to attract new customers and expand revenues by tapping into the massive and growing cloud connectivity market.
After a sunny week in San Francisco at this year’s Gigaom Structure event, a few consistent themes came through. One being security, or rather trust. Following a year of Snowden, Heartbleed, Target etc…, it’s clear that trust is a fundamental pre-requisite of being in the cloud business for all of us – from providers to infrastructure vendors. But trust still won’t persuade top retail chains like Walmart to move their cloud business to the likes of AWS, opening up opportunities for fresh, young, nimble cloud providers like Joyent to carve out a nice business. (And yes, Bryan Cantrill, CTO of Joyent, was a wonderful, charismatic and entertaining presenter).
Despite the security theme, several large enterprise organizations, including GE, discussed plans to rid themselves of all of their data centers (in 3 years) and in one session, Florida Crystals (you may know them by their “sugary” brand names such as Domino) are already 100% in cloud today. But there was an underlying backlash on public cloud as the private vs. public debate continues.
Surprising was the showcase of hardware - really. Facebook presented slides showing their new physical “top of rack switch” hardware. Not that cloud providers necessarily want to be in that business, but when big vendors can’t keep up with requirements, they can’t and won’t wait. An opportunity for nimble, innovative network infrastructure companies (yes, like BTI) to build what cloud providers want and need, versus force-fitting legacy equipment built for other applications.
Following BTI’s quick 2-minute “advertorial” strategically sandwiched between Microsoft and Twitter during the packed morning session on Wednesday, we moderated a fun, fast-moving showcase of 7 hand-picked young start-up cloud infrastructure companies during the LaunchPAD session – strategically timed before cocktail hour! The panel of judges – including Tod Nielson, CEO Heroku and BTI Systems board member, Michael Redding, Managing Director Accenture Technology Labs, and Samuel Angus, Partner Fenwick & West – delivered excellent feedback to each presenter and awarded CRATE Technology the esteemed Judge’s Choice Award (yes, I was given chocolates from the presenter, but that did not affect the voting in any way!). NIMBOXX was the winner of the Reader’s Choice Award following an enthusiastic group of supporters in the front row taking SMS voting to heart. All presenters did an excellent job under pressure to deliver their pitch in 4 minutes. You can view the Launchpad session online.
It was a packed schedule and as such I missed some great presentations (luckily all available online), including Intel, GE, Rackspace and others. But best quote (aside from the many from emcee Joe Weinman), was from DARPA presenter Dan Kaufman when asked if he had a Plan B – “No, because I always number my plans”!
Internet Exchanges (IX) and the Data Center Interconnect Market
As an ex-analyst cum marketer, I'm often asked how I would define and forecast the Data Center Interconnect (DCI) market. I've been somewhat surprised that no one has done this yet, as it would certainly include optical equipment, some core and edge routers and application appliances, resulting in a fairly large TAM - which we at BTI define as the Intelligent Cloud Connect (ICC) market. The ICC market has grown rapidly and significantly, in order to directly connect content to the end users or 'eyeballs' at the edge of our global networks. This shift of content closer to the eyeballs has resulted in the continued disintermediation of ISPs and increasing arguments over net neutrality and issues amidst paid and/or unpaid peering arrangements, which only begs the question; how is the content actually getting to all these end users, and is the byproduct of this shift additional data center interconnections in the metro?
The answer is clearly yes. The market has been growing on a global scale on national long haul routes, sub-terrestrial links and now massive and accelerating deployments in the metro regional areas. Driven early by the largest content providers (Web 2.0 companies), service providers are now building metro overlays for DCI, colocation and hosting providers are building DCIs as their businesses and real estate sites grow, and Internet Exchanges and CDN networks are helping to fill up these colocation and hosting data centers, adding massive amounts of content and thus more traffic.
An interesting and often overlooked aspect of the market is the Internet Exchange (IX) players. The IX operators differ from colocation and hosting providers in that they are largely non-profit (exception being DE-CIX), most are headquartered outside of North America, and they are entirely focused on getting content to their local users. The largest IXs in the market today are AMS-IX (Amsterdam), DE-CIX (Frankfurt), LINX (London) and Equinix (see Equinix Selects BTI Systems Solutions for Interconnecting Data Centers in Singapore). The number of IXs worldwide has increased by 20% since 2012, and the overall traffic on IX networks this last year increased 26%. In 2001 the Euro-IX was formed to combine resources, share information and establish standards of operation, and the resulting union has provided this type of traffic information that is largely unavailable anywhere else in the world.
The largest European IXs are beginning to deploy in North America in order to introduce this European model which is neutral and member-governed. LINX CEO, John Souter said "There has been a lot of attention worldwide for what we're building here. Our members endorsed the idea that LINX should build a neutral, member-governed and distributed Internet Exchange and that is precisely what we've done with LINX NoVA in North Virginia. The formation of Open-IX indicates that US operators are ready for a more transparent, mutual approach to peering, similar to what LINX has pioneered in the UK. This method of passing traffic is sometimes described as the 'European model' but it's primarily about putting the member first." Open-IX is an initiative started by a group of Internet network operators, content providers, and other market parties that encourage the development of a neutral and distributed Internet Exchange model in the US. AMS-IX in NYC received the first Open-IX certification for the New York/New Jersey metropolitan area, they also reside in Hong Kong. DE-CIX has UAE in Dubai and they are all making plans for additional sites.
The reason many of the content and CDN members are behind Open-IX is to break some of the higher prices set by some colocation providers in North America. Overall pricing for inter-exchange in Europe is significantly below North American pricing. However, another difference is that North American colocation providers typically deploy Layer 3 networks inside the exchange points, whereas most IXs provide a Layer 2 shared VLAN peering service and do not get involved in any Layer 3 routing.
Content providers and operators of CDN networks worldwide are encouraging service providers to connect to IXs in order to get content faster to all local networks. Since IXs and CDN players often reside in colocation data centers, the resulting number of colocation data center players and sites have increased dramatically, resulting in more data center interconnect in all regions of the world. IXs offer primarily public peering (and some private peering) services to hundreds of ISPs and content providers that span North America and Europe, however, areas like CALA and Africa are still vastly underserved, and because of that, traffic often has to flow from one area in CALA to Miami and back in order to reach other areas in CALA. Lack of local peering infrastructure in these areas as well as within China, is resulting in unfluctuating high bandwidth pricing in these emerging countries.
Because most (not all) IXs are transparent with the public peering traffic data, we get information about traffic patterns that we cannot see in our local North American markets. Clearly the local traffic is now staying more local according to EU-IX reports, and increasing at all sites, and very much so at the largest IXs, some of whom can peak over 1T. As more 'direct peering' takes place, the dis-intermediation of smaller ISPs is likely to continue, more data centers will be deployed, and growing content between the data centers is increasing traffic.
All of which is leading to the need for dynamic bandwidth on demand, and the ability to provision that bandwidth in minutes or seconds. Traditional and legacy data center interconnect deployments are simply 'glued' or 'static optical routes' for point to point transmission. This can no longer suffice; IX deployments require a new architecture - one that can collapse the layers and provide immediate analytics for network optimization - one that we call the Intelligent Cloud Connect. The Intelligent Cloud Connect architecture enables content, colocation and IX players to optimize their networks for faster service creation, faster content delivery and immediate end user satisfaction because it combines the need for massive scale with the intelligence of routing. The Intelligent Cloud Connect also enables SDN orchestration and open APIs for NFV-like services which enable increased monetization of the networks we are building today. A software control mechanism like SDN could interface to the IX infrastructure and then dynamically create Layer 2 or Layer 1 interconnects between the members. It could also provide a configuration portal for members. Other useful tools could assist in moving a public peering connection onto an IX hosted private session, which might provide increased security. A number of IXs have recently suffered from things such as route server and DNS-DDOS attacks. While most IXs today will remain non-profit and focused on serving their members quickly with SLA self-service portals and connections to multiple IaaS and content providers, they too, can benefit from the Intelligent Cloud Connect. Only BTI Systems enables content, colocation, Internet Exchanges and service providers to build better and more competitive data center interconnects with a new category of solution - Intelligent Cloud Connect. In Part-2 we'll discuss how Netflix and other CDN networks are also contributing to this market.
It was both enlightening and encouraging to hear the positive response to our recent announcement describing our expanded service and support capabilities - highlighting the importance to our customers of not just good technology and product, but also the extensive technical service offerings that go with them, particularly in this era of "software defined everything".
In a time when cloud and content providers are capitalizing on the many business advantages of building their own networks to deliver high-capacity connections between cloud data centers, access networks, peering points and partner networks - advantages that include not just cost but also control - leveraging equipment suppliers as business partners to monitor, manage and optimize these assets speeds time to deployment, optimizes operational resources and enables them to monetize their investment more quickly.
Case in point is Rackspace Hosting (NYSE:RAX). Rackspace chose BTI's technology solutions to enable simple, flexible and high capacity interconnectivity between data centers, colocation centers, peering points and POPs in key metro areas. In addition, Rackspace signed up for BTI's Managed Network Operations Center (NOC) and technician services to augment and streamline its own operations, and provide continuous 24x7x365 remote monitoring and management. The result was extremely quick time to deployment and ongoing SLA guarantees.
"Our Dynamic Optical Layer network was up and running within a couple of months of our first meeting with BTI. Working closely with our engineers, BTI ensured the deployment went smoothly and continues to do so", said Chip Hanson, network engineering manager at Rackspace.
And even traditional service providers can benefit from augmented monitoring, management, provisioning and technical support to ensure efficient and continual operations. In fact in a recent survey, 97% of BTI's service and content provider customers found BTI's technical service and support much better, better or at least as good as other vendors they have used.
So as service delivery continues to expand and as speed to deployment and guaranteed service levels becomes even more critical, network operators of all stripes will turn to their suppliers for extended support capabilities, and it is incumbent on those suppliers - large and small - to demonstrate this expertise, capability and knowledge on a global scale.
For nine years as an analyst I wrote reports on OFC, so it seems a bit weird this year not to write one. On the other hand, I had far less time to see panels and presentations as my 'new' job (meeting with customers, talking to analysts, and being on multiple panels myself) required every moment.
My takeaway after four days at OFC were:
- The market in general is relatively upbeat and OFC was well attended
- Component vendors are not playing chicken and egg anymore (i.e. how much volume will you buy and then maybe I'll make it), and have an increased urgency to deliver
- The lines between content providers and carriers continues to blur
- 25% of the content overall was SDN related
- SDN related material was more about implementation than explanation, and focused on faster service creation
- Technical sessions were mostly standing room only
- Standards are way behind implementation and business requirements
- Carriers this year are clearly making changes faster than ever before
- Churn within vendors (layoffs or internal re-orgs) was prevalent everywhere
- Carriers and content providers alike are demanding more from OEMs in terms of new solutions, more open solutions, and faster product delivery cycles
It was also exciting that BTI is becoming an interesting company to many who had not known much about us before. Even a large carrier said there was lots of "BTI buzz". There were still a few who had never heard of BTI and that surprised me. So, I encourage you to go to www.btisystems.com/resources page and read white papers and solutions briefs as we are very transparent about our product and company. Even our data sheets are useful, and we have a cool new video that really shows how we use the expressway for optical traffic, highway IP traffic in an intelligent manner and link in the serviceway when appropriate to tie applications directly to the network.
And, this week ACG Research released their TCO and Business Case for BTI's 7814/7802 packet optical and SDN/NFV based platform that clearly shows the extraordinary savings when doing 'real' or third generation packet optical transport that envelops and enables application awareness with an IT server model. To quote Michael Kennedy from ACG: "BTI is the only vendor offering LSR functionality on a packet optical transport platform". Hopefully, with these links if you don't know about BTI, you can learn more, and if you do know us, follow us on LinkedIn, Twitter and Facebook to get all our blog updates.
We recently became members of the European Telecommunications Standards Institute's (ETSI) Network Functions Virtualization (NFV) industry specification group. We haven't been shy in our support of openness, standards and interoperability, having also recently become an active participant in the Open Networking Foundation (ONF). Both of these initiatives reflect our strategy and vision towards helping network operators control costs and accelerate revenue growth.
NFV is important for operators because it expedites services innovation and velocity, and controls costs by leveraging standard IT virtualization and orchestration techniques to consolidate network equipment functions onto commercially-off-the-shelf (COTS) compute. NFV offers operators the opportunity to move away from the forced-fit legacy middle-box approach to adding application awareness, by decoupling software from hardware. This fuels much faster rates of innovation and new services, and promises to enable operators to drive higher margins through better network utilization and innovative service monetization with greater agility.
ETSI NFV is particularly pertinent for us as this initiative aligns with the announcement last year of our new category-defining platform - Intelligent Cloud Connect. An SDN-enabled, converged networking platform, Intelligent Cloud Connect integrates an easy to deploy Linux-based COTS platform for fast and flexible implementation of NFV-based applications, and our participation in ETSI NFV will ensure close alignment with other leading companies in the roll out of these transforming initiatives.
Stay tuned for more ........ !
A great deal of buzz at the Open Networking Summit (ONS) this week with excellent keynotes from Facebook, Google, Microsoft and more. But probably one of the most notable was John Donovan, Senior VP, AT&T Technology & Network Operations discussing how the company is transforming its infrastructure, its processes, its partners and its culture and leveraging SDN/NFV for the new "User-Defined Network Cloud". You can watch it here:
This was refreshing, especially having listened to one speaker - just fresh off the plane from Barcelona and Mobile World Congress - remarking how during MWC week, hotels, restaurants, taxis, flights and everything in and around Barcelona increase X-fold for the event. But...communication services don't - they stay at the same rate!
Carriers missed out on a huge opportunity to not only change prices, but more importantly, to offer value-added services during the week - such as guaranteed bandwidth, or drive targeted advertising to the well-profiled attendees.
Hopefully this is a wake-up call for the need to change, and the benefits that SDN/NFV will bring to enable that change for carriers, like AT&T, looking to transform themselves in this new era of communication services.
Just released - a new video describing how BTI combines application-awareness with massive scale and capacity in an open and programmable converged networking platform purpose-built for cloud networking -Intelligent Cloud Connect. Uniquely combining three levels of functionality: the ExpressWay, the HighWay and the ServiceWay, the video shows how BTI enables new and innovative levels of services and applications delivery for content, colocation and service providers looking for a new approach to cloud data center connectivity. Watch it now!
Executive Viewpoint 2014 Prediction: BTI Systems - SDN and NFV Drive Data Center Virtualization in 2014/2015
Virtual-Strategy Magazine recently published an Executive Viewpoint article by Sally Bament on what is driving SDN and NFV virtualization in the data center.
Lee Doyle, Principal Analyst for Doyle Research, recently published a vendor spotlight on BTI Systems. Click to download the report: BTI Systems Vendor Spotlight: Delivering Innovation to Packet Optical Convergence.
Applications and Network Integration - A Monetization Rant
Eve Griliches goes on a rant about monetizing the network with applications and network integration. Get into the conversation by visiting SDN Central.
BTI Systems 2014 Predictions - Virtualization to Transform Network Infrastructures in 2014/2015
VMblog.com recently published a blog on BTI's predictions for 2014.
I had the opportunity to attend the Carrier Network Virtualization conference last week, and had the privilege to learn what's going on in the minds of many carriers regarding SDN and NFV. Carriers are realizing that they are going through an inflection point - and one that doesn't come often - where SDN and NFV are true forces to be reckoned with.
Carriers realize that SDN and NFV definitions, concepts, business models, and solutions are too cloudy to understand. This is usually a characteristic of the early phase of any transformative technology. One presenter placed a dollar value on the status quo - the current $77 billion dollar network infrastructure industry. That made me reflect on another reason there is confusion in the market place. SDN-NFV technologies will have a lasting impact on how the current industry transforms, especially if it makes or breaks incumbent vendors. That puts a huge amount of inertia behind any change!
Carriers are by nature very conservative in adopting new technology. It comes from their underlying business model. By contrast, the OTT provider's business model is to offer an application experience for free to end users and capture value through targeted advertisements. In this model, the end users are more tolerant of any degradation in experience due to infrastructure failure or congestion. But in the case of carriers, the main revenue source is the end user and as such, they don't have tolerance for any experience degradation. End users are used to five 9s of high availability and they will continue to demand that. Carriers also realize they can't use that as an excuse to fall behind the innovation curve of OTT providers.
So, what carriers want to know is what benefits does SDN-NFV bring them? Is it CAPEX savings, OPEX savings or monetization opportunities? Carriers are worried about the challenges SDN-NFV will bring to their network operation model, OSS/BSS integration model, purchasing model and overall carrier culture and organizational change. Unless vendor's SDN-NFV solutions offer clear compelling value propositions, you won't see carriers embracing the challenge that comes with adoption. At present, OPEX contributes a major part of a carrier's cost of delivering services, so most are focused on optimizing OPEX rather than CAPEX. But they are also not sitting idly by; many have tasked their R&D divisions to "kick-the-tires" of SDN-NFV architectures and get involved with proof-of-concept validations of vendor solutions. But, adoption will be slow and the migration strategy of any solution will be key.
At the conference, I participated in a focus group to discuss the very question: "Is it CAPEX, OPEX or Monetization?" I was taken aback by some carrier's mindset of thinking that it is a zero-sum game. They worry that vendors will simply transfer margins from hardware to software. They believe that this approach won't drive down the cost-per-bit to contain margin pressure. They seem to have constrained themselves within CAPEX-OPEX optimization. So what about "Monetization"? Yes, it is not easy and innovation from both the vendor and carrier community is required. Vendors can start by building a new breed of intelligent networking platforms as building blocks enabling a monetizeable network fabric. Carriers can start by enabling Virtual Network Function (VNF)-as-a-service as a first step in the path towards monetization. If we just focus on driving out margins of both vendors and carriers, we should be mindful that we will drive out the innovation and will turn networks into a pure utility. So, is it OPEX, CAPEX or Monetization? Let me know your opinion.
The cloud is fundamentally changing the way applications and services are delivered and dramatically impacting the way networks are engineered and managed. Cloud, content, colocation, hosting, and service providers all require high capacity, resilient, low-latency backbone networks to enable data center virtualization, workload mobility, business continuity and disaster recovery, and tenant-to-tenant connectivity. Legacy networking solutions designed to support traditional enterprise applications and data transport services are too costly, complex, and inflexible for today's dynamic cloud environments. Conventional metro-regional backbone networks are constructed using separate layers of discrete switching, routing, and optical networking elements that support specific protocols and functions. Each network layer is provisioned and managed separately using distinct administrative interfaces - a costly and time-consuming process.
By delivering a purpose-built networking platform with integrated packet-optical hardware and highly optimized software, BTI has created a truly disruptive platform that enables unprecedented economics for cloud providers. Providers for the first time can take a top-down, applications and services perspective of the network - shaping traffic flows and adapting bandwidth across the extended network fabric in real-time to address fluctuating application demands. ICC provides an intelligent, abstracted view of the network rather than focusing on lower-layer functions. Services are provisioned holistically across the fabric, rather than individually, node-by-node. LSPs can be established and added dynamically to traffic engineer across the network fabric - for instance, to enable virtual machine mobility across virtual data centers or to allow a key customer to request bandwidth in real-time.
The BTI Intelligent Cloud Connect - the industry's most scalable, cost-effective and extensible inter-data center connectivity solution - is fundamentally transforming the economics of cloud connectivity.
Download the white paper here.
Part 1: Multi-layer SDN Control
It is invigorating to see the flurry of recent announcements such as Cisco's ACI, Juniper's MetaFabric and Arista's single-tier Spline architectures for network virtualization and application-centric intra-data center architectures to drive simplification. Cisco and Juniper have been emphasizing the importance of having the intelligent infrastructure layer along with the global view of centralized control functions within an SDN context. The main narrative has been "centralize what you can and distribute what you must". For example, we have to see data center incumbent Cisco's ACI architecture within this narrative to watch how the Nexus 9000 Series fulfills the requirement of the intelligent infrastructure (distribute what you must) and how the Application Policy Infrastructure Controller (APIC) fulfills the centralized network virtualization (centralize what you can) through a policy-centric application integration point.
This network virtualization, centralized policy-centric application integration and simplification of underlying networks are critical aspects of bringing out the SDN value proposition of an elastic network layer that responds to the dynamism of emerging applications. But, as the multi-data center model becomes the modus operandi for major content, colo, hosting and cloud providers, the SDN value proposition needs to be extended beyond the physical four walls of the data center and extended below the routing and switching layers. Currently the static, overprovisioned optical layer that interconnects highly interactive and chatty data centers limits the SDN value proposition to application-centric architectures within the walls of the data center, if the focus remains within those very walls.
Many operators want to move into a "pay-as-you-go" model and virtualize the bandwidth as an elastic resource pool - not only to monetize but also to increase the utilization of the underutilized and overprovisioned bandwidth capacity. These operators want to give the control of "bandwidth-boost" directly into the hands of their tenants, to manage and control these elastic resources. In this context, elastic network virtualization needs to be extended to the exponentially growing optical bandwidth resources that interconnect these highly chatty data centers in highly-populated metro and regional networks.
Intra-data center network dynamism, programmability and automation is the first step, but it can't stop within the four walls of the modern data center. The SDN value proposition must be extended to lower optical layers that interconnect the modern data centers in metro and regional deployments, to unshackle the network in an end-to-end manner and enable network optimization and monetization at a whole different level.
Part 2: Application-aware Intelligent Platforms (coming soon)
I have always said BTI is a unique company - large and experienced enough to deploy and support large-scale global networks, and yet nimble and lean enough to innovate.
And that was recognized recently by the IEEE who nominated and selected BTI for the 2013 Outstanding Technology Recognition Award in Information and Communications Technology. As a result of a combination of our continued investment in the research and development of advanced networking infrastructure solutions; our success in delivering solutions to leading content, colocation and service providers around the globe; our active role in global industry standards bodies including the ITU-T, IEEE and MEF; as well as our international partnerships, BTI was selected "for continuing to push the envelop in the development of high-density optical networking for content service providers".
We thank the IEEE Ottawa Section for this prestigious recognition, and are proud of our continued leadership and innovation in networking.
Want to read the press release: "BTI Systems Receives IEEE Ottawa Section 2013 Outstanding Technology Recognition Award"
Sally Bament, SVP Global Marketing, talks about how BTI deploys SDN and NFV and is open and agnostic to controllers.
Hear from Paul Crann, SVP Product Management, on why BTI's Intelligent Cloud Connect platform redefines cloud networking by combining network intelligence and application awareness with massive scale.
See why BTI executives and market investors believe open platforms are the heart of next generation network architectures.
Posted in Business and the Network, Cloud, Data Center, Industry Insights, Intelligent Cloud Connect, NFV, SDN, Video | Tagged Data Center Virtualization, Intelligent Cloud Connect, SDN, YouTube | Leave a comment
More than a year after Facebook's IPO, we are seeing other social networking sites appear and expand dramatically - particularly in regions of the world where there is a need for native-speaking users to stay in touch with one another.
Case in point is VKontakte (VK) - now the largest social network in Europe with over 100 million active users around the world. And according to a recent New York Times article, some investors are funneling proceeds from Facebook IPO into VK.
And BTI is part of VK's expansion plans - interconnecting data centers in St. Petersburg and points of presence in Moscow, Frankfurt and Stockholm. For more information on this deployment read the press release.
Q: How does BTI's focus on SDN and NFV help your customers?
A: Our focus is mainly on cloud providers, colocation/hosting providers and Web 2.0 customers who are early adopters of technology and architectural changes. These are the providers who clearly 'get it', and are on the fore front of building high-capacity, low-latency scalable networks that demand high performance. By adopting the latest technology, networks can become quick and agile and can deliver services at a very faster rate. This agility fits the new model for NFV that delivers bandwidth or services in 'software time' not the standard hardware time that was 18 to 24 months.
Q: What are some use cases you are seeing?
A: The typical use case being discussed today is virtualization, but we see many more beyond that. Examples are server extension to access networks from the POP, and load balancing across multiple data centers. A real interesting one is where hosting and colocation properties provide not just traditional interconnectivity, but an actual partner and customer portal for meet-and-greet business opportunities. This enables a whole secondary source of income for the provider. What's key to these applications or use cases is that they require an open networking platform. And, an open networking platform is extremely useful for the architectures that cloud providers, content delivery networks and private backbone networks require to deliver new services.
Q: How will those use cases be introduced into the network?
A: Well, it's certainly a process. Everyone, whether they're an enterprise, a service provider, or a Web 2.0 company, is looking to either make money from their network, save money on their network or be a differentiator one way or another. SDN has made everybody rethink 20 years of networking and say let's do this a little bit differently, which is great. Before, no one really knew if the applications were impacting the network. Was it the application, or was it the network? We're at the point now where we can pull out the data, do a little data mining, and as one customer said, look at 'empirical evidence' not someone's guess on what the application might be doing to the network. Once this happens, we can program and automate the network so it's tuned to what the applications require. Those two processes really need to take place immediately.
The next steps will be to offer more services, and offer those services in a faster time to market. Traditionally, there has been an 18-to-24-month delivery for a new service to market. That's unacceptable these days. We're now talking about weeks or days to get new services to market. That is really where the monetization comes in; faster time to market brings revenues in sooner. Once that is in place, providers can do more bundling and subsequently more upselling. And having said that, there's actually a secondary monetization market which I believe is still relatively ignored. 75% of the Netflix downloads today come from their online recommendations. Look at Zappos and Amazon, all of these sites operate and have recognized increasing revenues due to their recommendation models. So, multiple monetization models are beginning to emerge and that is what providers are discovering. I think this is where the exciting applications and market are going to be.
Q: What are some of the inhibitors to SDN and NFV adoption?
A: It's not the technology. But clearly, smaller companies are more nimble than larger ones, so there will always be an advantage when change is rapidly taking over a market like this one is right now. Some of our customers have assembled their most innovative staff, created an internal prototype and presented it to the rest of their company, with the intent that others will join the team having seen what can be done.
Q: Can you give an example of a customer utilizing SDN/NFV today and how?A: We already have customers that have connected large data centers, moved their core routers to the peering or gateway function, and started to switch instead of route between data centers. They are using our open networking platform and applications enablement platform to bring in new customers who can partner with other customers in their network. Hosting and colocation providers are already enabling their customers to meet other customers and do business with them. This is enabling a whole monetization neighborhood out of a simpler interconnectivity model, and it's all being done through open APIs. With an open networking platform that can enable cloud to cloud, cloud to colocation host, hosting provider to customer, and customer to customer interactions, a new revenue monetization market can be created on top of simple and easy infrastructure.
We are extremely excited to share that Digital Realty Trust, one of the world's largest data center providers, this morning announced that they are launching a new cloud data center connectivity initiative that interconnects major data centers in the London metro area. The service will also provide connectivity to key London metro gateways and over 500 carrier, service, internet exchange, OTT, cloud, content and local access providers around the world.
Leveraging our innovative Cloud Networking Solutions and Epsilon's Global Network Exchange, Digital Realty Trust will create an intelligent application-aware network fabric targeted at cloud and content provider customers. The announcement represents so many exciting things, but most importantly, it represents a first-of-its-kind cloud-centric collaboration leveraging technology and solutions from three innovative market leaders - Digital Realty Trust, Epsilon and of course, BTI Systems!! Exciting times! Read the announcement here.
Every five years or so the trend to solve the age-old question about how to build networks that are simpler and easier to use comes around again. Some have had traction, others fell to the wayside (remember PBT?). The buzz around software-defined networking (SDN) and network functions virtualization (NFV) makes me wonder if we are reliving this again, or if there is more substance behind the movement this time? The answer is yes, there is more behind it this time. The reality is that networks simply cannot continue to be built the same way because we have hit major scaling issues. Traffic growth is on an order of magnitude that network hardware simply cannot keep pace unless it's repurposed for a single function - just to be fast.
There are really two movements afoot. The first is to get the hardware positioned to perform at its peak, the other is to orchestrate increasingly intelligent software in a way that does not behave like older protocols. Note the advent of white boxes sold to date and you know things are rapidly changing. So, what do the white boxes do that network infrastructure couldn't do? They are by nature open platforms for new operating systems and new control mechanisms. Focused mainly within the data center they spark the need for open networking platforms in the metro, regional or wide area. Wide area networking is evolving to use simple and fast hardware where needed, yet have the intelligence where needed as well. In fact, an open networking platform can be the basis of a complete server approach to integrating analytics and application awareness, thereby launching a platform that can thrive within the NFV environment.
What benefits come from this new environment and why it is different this time, is because the change enables significant improvements that enable monetization of the network. Having stated this, we still have a few stages to proceed through, one which we are on the precipice of, is mining the network with advanced analytics to understand how the applications are running and to make changes to support the applications. Then we can automate that process and begin the programming of the network process that will automate how the network will adapt to the applications. Once these are in place, the network has been established to deliver services faster (remember fast?) and deliver increasing numbers of services in parallel or overlapping. Service creation time that is shortened from 18-24 months to 6 months or weeks or even days - changes the playing field dramatically, enabling a cloud or service provider to operate in 'software time' rather than hardware time.
But wait, there's more. It's really the secondary form of monetizing the network that has the largest potential, and one that I don't think is understood. That secondary monetization uses all the previous steps to enable information from the network to deliver recommendations and advertising which can be targeted for greater value and generous rewards. Because it's not just about taking cost out of the network, it's about creating a tool to monetize the network, and to differentiate your business.
We attended GigaOM Structure in San Francisco this past June. Here are some live shots from our first day – conversations revolved around who attended and what the ecosystem was like, while in the background, the BTI band (The Dealmakers) tuned up. More to come soon!
So many of us in the networking and tech sectors think that the whole world lives in our "pool" and understands all of the trends, acronyms and technology alike. So while we are working hard to move the needle - particularly when it involves technology that in actuality most everyone uses everyday - we sometimes need a reminder to speak in terminology that we all understand. Case in point is an article sent to me recently (dated from last year)......
Ray Mota, PhD and Managing Partner at ACG Research (www.acgresearch.net) speaks out on the economic landscape, the benefits and momentum of BTI Systems, and his ‘baseball’ relationship with Eve Griliches.
Sally Bament - named to Fierce Telecom's prestigious list of most influential female executives, Women in Wireline 2013
SVP, Global Marketing
Just this week our SVP of Global Marketing at BTI learned she had been named to a list of influential female executives by FierceTelecom. I was thinking about how we could share Sally's acknowledgement with our community and volunteered to write an article for BTI's Blog. No sooner did I start to pen the article, when I received an email from Steve Waszak, President and CEO of BTI. After reading his message, I realized my work was done. I couldn't think of a better way of saying what he thoughtfully shared with all BTI employees:
A quick note to bring your attention to an accolade earned by Sally Bament. She has been named to FierceTelecom's prestigious list of most influential female executives, Women in Wireline 2013. I've included the link below for more information.
One of 10, chosen from hundreds of candidates identified by the editors of this top industry publication as well as other influencers and insiders, Sally joins fellow winners from such titans as Google, AT&T, Verizon and CenturyLink being lauded for 'shaping telecom services, innovation and policy'.
Anyone who has spent even five minutes with Sally immediately recognizes that her tremendous passion for moving the industry forward is matched by her unparalleled work ethic in turning vision into reality. The pivotal role she has played in spearheading BTI's rebranding -- notably, her leadership in helping raise BTI's visibility across leading analysts and our customers -- is emblematic of the tireless dedication that has characterized Sally's career. It's why she's respected by colleagues, customers, media, analysts, partners and peers.
We're incredibly fortunate to have Sally as part of our team. Please join me in congratulating her on this well-earned honor -- and thanking her for embodying everything BTI stands for.
Steve Waszak | President & CEO BTI Systems"
Earlier this morning Chandra Pandey, BTI's VP of Platform Solutions and Lane Patterson, CTO, Equinix spoke at GigaOM Structure on the topic, "Compute Everywhere: The Architects Behind The Change". The session focused on the demands posed by data center virtualization and the need for more distributed yet faster and more capable infrastructure. Watch the session at GigaOM's event archive, or by simply clicking on the embedded video below.
I spoke with many service providers, content providers and enterprises about the growth of the data center interconnect market and how it has impacted their business, and found that their focus was on next generation network architectures. Metro traffic flows are straight forward, yet there was a lot of talk about using VPLS, L2 VPNs and L3 VPNs to control the traffic. I looked at that and asked myself, why are we adding all that overhead for what is essentially machine to machine traffic that is just going to peering points or gateways? That challenged me to think differently.
For content-centric providers, it was all about scale and cost, because content providers don't see the network as a service, they see the network as a vehicle to deliver their applications. I also talked to traditional service providers and they were on the exact same page. The difference is the transition for service providers will be longer, but the ideas are the same. It was clear that the evolution of the network had to be scalable and cost-competitive for both markets.
No one gave us the architecture, but it was clear that we had to collapse the network layers and be competitive in scale and convergence in order to provide a path on which to evolve. Because you never know when and where the next disruption will occur, it was key to plan for an open networking platform that would provide a framework for innovation. We also realized that traditional vendors did not have the ability to innovate and create a solution to address this.
The next step was to bring in application awareness. Traditional service providers look at individual users or the enterprise, but a content provider looks into the applications, because they care about how that application is performing in the network. Being able to determine the exact resources an application requires is critical. Only an open networking platform can do this. We developed a platform and a framework, and that framework allows you to write applications or insert an application and decide where you want to run it. We validated this architecture many times, and our customers became excited that it could be done and scale within their networks. Others have tried this, but have not achieved proper scale. We decided to bring the performance directly to the application.
Our customers are very excited and we are excited. We have the open platform, the application interface, and we are running applications today. We can add a new application in days, no one else can do that. We call it in 'software time'.
I was just thinking back to my analyst days (ok, only a few months ago) where I would go for information on a new market. Of course searching the web was where I started, but for new technologies and markets such as SDN and NFV, where is there to go? I found a rich set of bloggers who consistently kept reality firmly in tow and did not tend to bend with the market hype. As the Solutions Marketing lead at BTI, I'm responsible for our blog content, ensuring that it provides value and thought leadership for our followers. With that purpose in mind, I first went to collect the bloggers or blog sites that have impressed me the most to date. I'm sharing a handful of blogs to highlight who has led the way, and will continue to add to our 'Blogroll' over time. With that, here's who they are and why.....
First, the irascible Tom Nolle has been very vocal for his clear delineation of each vendor's SDN and NFV (or lack thereof) strategies, poking holes yet highlighting successes. While clearly an analyst, Tom does not hold back on any vendor. Tom speaks his mind, and with him we all learn. You can find his blog at http://blog.cimicorp.com/.
When people ask me who the SDN vendors are, and where do they play I go to SDNCentral http://www.sdncentral.com/sdncentral-blog/. More than a blog site, SDNCentral posts daily comings (not too many goings) of SDN companies and categorizes them into a tangible market to understand.
GigaOM.com could be considered many things, analyst site, blog site, technical information site, all of which is true. While our market shifts to 'software time' GigaOM has successfully encompassed technology reports from optical to the heaping family of software companies, all of which are written with an open attitude and intention to educate. GigaOM is at http://gigaom.com/ and has an SDN site as well.
And last, but by no means least, is Data Center Knowledge at http://www.datacenterknowledge.com/. Skilled reporters cover data center build outs, expansions and always seem to have a hot thumb on who's doing what in the data center area. Very helpful for daily news on the industry.
Ok, so those are my top picks for this week....if you feel I should be reading your blog too, please send the link to firstname.lastname@example.org and I'll give it a look. Until then, please follow BTI Systems on LinkedIn and Twitter (@BTISystems) as we continue to reveal new customers, new products and new ways of looking at the network. Let's all stay connected.
BTI has been busy preparing for some great events this month, starting this week with NANOG058 in New Orleans on June 3-5. Then we head to Singapore for CommunicAsia 2013 on June 18-21, and San Francisco for the GigaOM Structure event on June 19-20. BTI's Chandra Pandey will be speaking at GigaOM on the topic, "Compute Everywhere: The Architects Behind The Change", alongside Lane Patterson, CTO, Equinix. The session will dive into the demands posed by data center virtualization and the need for more distributed yet faster and more capable infrastructure. And if you arrive Wednesday, BTI is sponsoring the first night's reception featuring our very own rock band The Dealmakers, starting at 6pm. Register now so you don't miss the brilliant lineup of speakers and topics. We hope to see you in SF!
Many content providers (CPs) have discovered the advantages and major cost savings of building and managing specific sites within their own networks versus leasing managed service wavelengths.
Because key points in the network are now requiring cost evaluation due to bandwidth growth, CPs are using a TCO model that compares outsourcing with a build-it-yourself approach to evaluate each route installed. Based on these inflection points, CPs can start in one metro area, experience the benefits, and evaluate the next area of their network for optimization.
Managed wavelengths are expensive and have hidden costs such as IRU taxes. As additional bandwidth is deployed, the additional cost is linear, often becoming untenable at certain breaking points. In addition, slow services and lack of urgency from managed service providers leave CPs lacking control over their own network, especially when faced with unplanned bandwidth upgrades.
Benefits and Drawbacks of a Managed Service
The benefits and cost savings associated with branding, legal protection, security, expertise in technology, global connectivity options and service choices - network-related benefits which up until now were beyond the budgets of companies - contributed to the viability and affordability of managed services, making it a logical delivery option for content providers. A CP deploying additional bandwidth will find their monthly payments increasing dramatically with managed services. These issues alone are enough for most CPs to take the next step and examine the business case for moving from an outsourced model to building it themselves. But with the availability of affordable protection and restoration services as well as the low cost and savings of leasing dark fiber, especially in the metro, companies are increasingly bypassing managed services and instead building and managing their own networks between data centers, peering points and POPs.
Benefits and Drawbacks of a Build it Yourself Network
For CPs, the private network solution enables them control over network upgrades, expansions and extensions at a much lower cost point. Building the network themselves puts the control of personnel, resource allocation, costs, capacity issues and decisions back into the hands of the company. Content providers can utilize breaking point metrics for their high bandwidth routes and easily deploy and upgrade with security and latency firmly in their control. And although there are challenges with deploying your own network such as negotiation of RTU spots, inefficient use of equipment if bandwidth is low, your personnel and networking group are now making the real-time decisions related to your infrastructure and network.
The Business Case Analysis - a Total Cost of Ownership
The tipping point decision criteria to build versus outsourcing a network is based on dark fiber equipment and regional wavelength costs. The CP must evaluate the economics of each route, including current and forecasted bandwidth, in a price matrix and compare this to the managed service offering. When the TCO is complete, often major cost savings become apparent and are very compelling. To date several BTI customers have found at least 60% savings in a three year period by moving from a managed service to deploying their own rapidly growing network. Often, payback is within the first year or two, and first year costs are lower than the original managed service offering.
BTI provides unique, responsive and flexible solutions for your network building requirements.
BTI's Intelligent Cloud Connect integrates massively scalable packet and optical features into an open platform that seamlessly integrates network analytics and application awareness.
For more information on the BTI Build vs. Buy TCO contact: email@example.com
By Dr. Robert Keys, Chief Technology Officer, BTI Systems
Content and service providers are facing challenges as they scale their networks to support daily demands from consumers and businesses. Complicating these demands is an impatient public - for example, a 100ms delay costs Amazon 1 percent in sales and Bing found that a 2 second slowdown reduced queries by 1.8% and revenue per user by 4.3%.
The challenges they face are primarily:
- Delivery of timely and differentiated cloud services
- Better utilization of expensive resources in the network
- Reduction of OPEX and CAPEX costs
- Keeping up with scale and capacity demands efficiently
Areas that can most impact and improve performance and scale in the cloud - and are correspondingly ripe for improvement driven by new software and system innovations - include:
- Increasing application awareness
- Improving network analytics
- Automating control of the network by leveraging SDN approaches
- Reducing layers of complexity
- Driving scale and performance
Large content providers are already tackling these issues by leveraging flexible SDN-enabled platforms that deliver network intelligence, analytics and massive scale, density and increased performance, at significantly lower OPEX and CAPEX costs.
There has never been a better time for content and service providers to assess their cloud network opportunities and deploy new solutions and strategies to optimize incoming revenue streams. For more on this subject, see the article in Lightwave Magazine.
Eve Griliches and Sally Bament, BTI Systems, discuss BTI's new Intelligent Cloud Connect and how it enables content, cloud and service providers to overcome the performance bottleneck created by the exponential growth in traffic flowing in and out of data centers as a result of the dramatic demand for cloud services. An open, SDN-enabled networking platform combining routing, optical and applications intelligence, Intelligent Cloud Connect is designed to meet the stringent performance, scalability, operational efficiency and service innovation demands of the cloud.
Deloitte’s Tech Fast 50 list included Kanata-based middleware provider Solace Systems, which just missed the top 10 with a ranking of 11 and revenue growth of 1,512 per cent over the past five years. BTI Systems, Epiphan Systems Inc. and Optelian also made the top-50 list.
Cleveland, OH, October 9, 2012-Rock and roll bands comprised of employees from eight national companies owned the night at the 12th annual FORTUNE Battle of the Corporate Bands on October 6 where "The Grove Valve Orchestra" took home the title of "Best Corporate Band in North America." From center stage at the Rock and Roll Hall of Fame and Museum in Cleveland, Ohio, the bands played their hearts out for charity, with event proceeds benefitting the Rock Hall's educational activities. Read More.
- Business and the Network
- Company News
- Data Center
- Industry Insights
- Intelligent Cloud Connect