It’s clearly evident from the evolution of technology that the “need for speed” seems to be deeply embedded in human nature. Reflecting back without going too far back in history, the horse and buggy was the main mode of transportation, unfortunately not fast enough. So we invented the locomotive, automobile, airplane, fax machine, e-mail, and mobile phones with text messaging among the hundreds of other inventions to fulfill our need to do things faster.
Being a networking guy, I might be biased, but I see networks as the new frontier for speed, especially now that we are a media/information driven society. It wasn’t long ago that a 10Mbps shared Ethernet LAN and 56kbps WAN links were considered fast (showing my age here). However, every time faster networking speeds were introduced, newer applications quickly consumed the capacity driving the need for even higher speeds.
Over the years we’ve seen Ethernet speeds increase in increments of 10x starting with 10Mbps to 100Mbps to 1GE and 10GE and now, we’re again at another speed inflection point -100Gigabit Ethernet! This week Cisco added to our 100GE router portfolio (CRS and ASR routers) with the announcement of a 100GE M2-Series module for the Cisco Nexus 7000 Series switches. Along with the 100GE module, we also announced a 40GE M2-Series module for the Nexus 7000 and a 40GE module for the Catalyst 6500.
The Nexus 7000 modules deliver feature rich, non-blocking Layer 2 and Layer 3 functionality. The 2-port 100 GE module enables up to 32 non-blocking 100 Gigabit ports in a single Nexus 7018 chassis, while the 6-port 40 GE module enables up to 96 non-blocking 40 Gigabit ports. These high speed modules allow up to 10 times faster connectivity and bandwidth over existing fiber infrastructure, resulting in more services and data delivery over the same network.
So, what are the new applications driving the need for 100G connectivity today? Organizations and Service providers are experiencing a data deluge brought on by the confluence of a number of growing trends including faster residential connectivity, video, cloud computing, virtualization/workload, VDI and the explosion of smart mobile devices with rich media applications. For the most part, 10GE is today’s work horse for network backbones in the campus and data center, however, 10GE is becoming a bottleneck at the network core of many organizations, especially between the aggregation and core layers in the data center.
Based on recent customer conversations, we see early adopters of the 100GE technology in multiple areas:
1. Enterprise Data Centers: Today’s data intensive trends such as, Big Data, video, cloud, VDI and faster virtualized servers are accelerating adoption of 10 GE at the server level. This in turn is placing a heavy load on the upstream links resulting in high oversubscription between the access/aggregation layers and even higher oversubscription between the aggregation/core layers. Both 40GE and 100GE will provide the additional bandwidth needed to help reduce the oversubscription and alleviate these growing bottlenecks.
2. Academia and Research -- Research networks are transporting Petabytes of data across academic institutions for collaborative scientific research. To accommodate this data deluge, research institutions are deploying global high-speed 100GE networks with 100GE drop offs to research labs and universities around the world.
3. Content Providers – Leading social, search and gaming sites such as Facebook, Google, Yahoo, Tencent in China, to name a few, are experiencing exponential growth in traffic. To keep up with this growth and continue to provide fast response times, content providers are moving towards 100GE not only in their internal network backbones, but also for their links to service providers.
4. Service Providers/IXPs -- The new trends and the data deluge brought on by “the internet of things” from the home, enterprise, mobile devices, content providers and cloud are placing enormous bandwidth demands at the core of service provider networks, driving them to deploy 100GE connectivity within their networks and as high bandwidth handoff to their customers and IXP internet peering .
Looking out over the next couple of years, I suspect 10GE will still be a dominant player for most campus and data center networks, however as 10GE moves closer to the network edge (access and server connectivity), I see 100GE and 40GE becoming the de-facto standard for backbone connectivity….that is until we get bored with 100GE and feel the need for more speed -- Terabit Ethernet? Stay tuned.
Everyone is talking about the Cloud but many companies are just beginning their journey and want a guide to ensure the quickest and most cost-effective path. Questions arise such as “can we build a cloud in-house?” which is kind of like taking a bus to your destination if you don’t have expertise at your company. Others ponder “should we hire a partner to advise us?” which is the equivalent to taking a jet because an expert will quickly fly you to the cloud, leveraging knowledge that has accumulated through experience. The jet will get you there a lot faster with less bumps and obstacles.
Companies like Discovery and Innovest Systems have reached their destination to the cloud with the help of the Savvis Symphony Cloud Solution. To securely keep up with massive growth while aligning costs to revenue opportunities, Innovest Systems has been utilizing Savvis cloud solutions for more than six years.
Savvis delivers a complex, fully managed infrastructure for Discovery Communications to support its real-time advertising deliverables, ad sales, scheduling, and programming for all its international operations.
Positioned as a leader in the Gartner Magic Quadrant for Public Cloud Infrastructure-as-a-Service (IaaS), Savvis delivers a leading-edge solution that results in costs savings for its customers.
“We help customers with cost-affordable, reliable IT infrastructure services so that they can focus on applications that are differentiating them in their markets,” says Bryan Doerr, Chief Technology Officer, Savvis.
The Savvis cloud solution offers enterprise customers unprecedented flexibility in controlling how much applications costs and how effectively those applications are delivered to end users.
Savvis started with an IP Next-Generation Network (IP NGN) that is world class in both its network architecture and service delivery capabilities. Building on this foundation, Savvis then turned its attention to the infrastructure side of the data center.
Its vision led Savvis to create a unique cloud-based service, offering enterprise-required services, not just compute virtualization. The solution is designed to meet the broadest range of global computing needs which allows Savvis customers to focus on innovation, competitive advantage, and growth by freeing up IT resources.
Cisco has been part of the Savvis journey to cloud. Cisco Services helped Savvis create their cloud offering, achieve their goals, and realize the full value of its investment in data center and cloud solutions. This included, but was not limited to, an 18 month roadmap aligned with Savvis’ technology and business goals.
Cisco has just released a customer case study (PDF) of CareCore National, a 1200 person healthcare insurance company located in South Carolina. CareCore’s use case demonstrates how Cisco’s Virtual Security Gateway (VSG) for the Nexus 1000V can be used in a virtualized data center to logically isolate virtual machines running on shared application servers to meet compliance requirements. Read More »
We invited Mike Taylor, Vice President of global infrastructure engineering and operations at Savvis, to provide his insight on Savvis’ journey to the cloud. Read below for what he has to say. Check out related blog for an additional perspective.
If you’ve been following the cloud services market, you’ve likely heard the term “enterprise cloud” proclaimed by various vendors. But really, what does that mean? How do you differentiate an enterprise cloud from a mass market option?
At Savvis, a CenturyLink company, we love talking about our enterprise cloud offerings and what distinguishes them from the mass market clouds that continue to flood the marketplace.
First, let me be clear: In some areas, enterprise and mass market clouds are the same. Benefits for both include flexibility, quick provisioning of compute power and a virtualized and scalable environment. However, it’s important to note that enterprise clouds also provide a range of security options, unprecedented speed-to-market and vastly improved collaboration between the end-user and the vendor.
Savvis’ enterprise cloud is a VMware-based service differentiated by an array of built-in security features, as well as many optional managed security capabilities. Savvis built its cloud solutions using the same trusted suppliers – including Cisco – used by enterprise customers in their own data centers. Our cloud services are divided into tiers, providing different levels of performance and availability for different types of application needs. These services are delivered in a multitenant way and can also be delivered as a single tenant.
So how do you realize the promise of enterprise cloud infrastructure? My colleague Steve Garrou, vice president of global solutions management at Savvis, recently shared on the Savvis blog a list of items that should be addressed when considering a move to enterprise cloud. Rather than reinvent the wheel, here are the items that Steve outlined:
Decide whether you are going to maintain two infrastructures or consolidate.
Understand what applications are currently running in the existing environment and expectations for moving certain solutions to the cloud.
Analyze the architecture of the application environments.
Determine how much capacity you need to run the applications; are the capacity requirements seasonal or variable?
Assess compliance and security requirements.
Years ago – before “enterprise cloud” was common terminology – Cisco and Savvis shared a vision for a cloud service that offered enterprise-required services, not simply compute virtualization. That vision became reality two years ago when we launched Savvis Symphony Virtual Private Data Center, one of the industry’s first enterprise-class, multi-tenet cloud solutions. A key element of the cloud architecture was the Cisco Unified Computing System.
Partnering with trusted companies like Cisco helps Savvis set the bar for enterprise cloud. I recently sat down with Cisco to talk about our collaboration. You can see the results of those conversations in the case study and video.
For anyone who has ventured to a tech conference, flown into an airport or even driven down CA highway 101 this past year, it’s clear that cloud is still top of mind for many technical and business decision makers. We believe this means that enterprises are no longer just talking the talk, but are looking deeper into their networking infrastructure to see if they are ready to meet the challenges of cloud, virtualization and workload mobility. At Cisco, it is our job to help build clouds that can handle elastic demand and efficiently use the networking infrastructure at both a virtual and physical level. This week, we are announcing several key upgrades to the Nexus 1000V family that bring scalability and cloud readiness to the network.
Cisco is announcing this week a new member of the Nexus 1000V virtualization infrastructure portfolio, the Nexus 1010-X virtual services appliance. The new Nexus 1010-X is an extended version of the existing Nexus 1010 appliance, and represents a larger, more scalable and cost-efficient configuration for larger data center deployments and cloud applications. What is a virtual services appliance and why should customers use it? The Cisco Nexus 1010 and 1010-X provide improved management, scalability and visibility in environments running the Nexus 1000V virtual switch and the VMware vSphere hypervisor. Read More »
Its hard to believe that almost an entire month has gone by since the beginning of the year. The year has been off to a great start for Cisco Unified Computing System (UCS) now serving 10,000 customers. Reaching the 10,000-customer milestone is an achievement for relatively new and innovative platform.
Generally asset management implies financial management but this discussion is focused on operational management of the data center components. Typically, in Data Centers, different teams manage servers, networks and storage. These teams have cursory knowledge of each other’s domains. This organizational structure hinders data centers from obtaining higher efficiencies and agility. Data Center Management tools that allow automated workflows with enforcement of policies set by domain experts reduce time needed to effect changes and hence increase agility. Unified server, network and storage infrastructures with proper management capabilities improve overall efficiency, reduce data center complexity and promote better resource utilization. With Unified infrastructures the server management teams can make informed decisions on application workload placement based on their visibility into the network setup and policies set by the Network domain experts. For example, a server administrator could place more sensitive applications on servers that are connected to very secure network segments, or place bandwidth hungry applications on network segments with spare capacity. If network managers need to move network segment capacity around they would need the equivalent of network hypervisors. These decisions which affect multiple domains could be manually executed or orchestrated with systems management tools. The crowning glory would be for the end customer of the IT service to request infrastructure services from a catalog and get access to it instantaneously. A Forrester Research paper that Cisco sponsored even shows a maturity model for service orchestration within a data center. Where do you think your organization is on this maturity model?
by John Rollason, Senior Manager Product, Solutions & Alliances EMEA, NetApp
For many years the server market was dominated by the likes of IBM, HP/Compaq, Fujitsu, Dell, Sun and characterised by small market share shifts. True the market changed as rack and blade servers became popular, but most of the players recognized the shift and adapted. Then Server Virtualisation technologies changed the market and Cisco disrupted it completely with the launch of the Cisco Unified Computing System (UCS) in 2009. Today Cisco’s vision for server virtualization has been proven successful with more than 10,000 UCS customers and 54 UCS world record results. Customers obviously see the advantage!
Just over a year ago NetApp and Cisco introduced FlexPod, a pre-designed, pre-tested and validated Data Centre cloud solution built on modular and unified architecture composed of Cisco UCS servers, Cisco Nexus switches, and NetApp unified storage systems running Data ONTAP. FlexPod components are integrated and standardized to help you eliminate the guesswork and achieve timely, repeatable, consistent deployments. FlexPod has also been optimized with a variety of mixed application workloads and design configurations in various environments such as virtual desktop infrastructure and secure multi-tenancy environments.
Today more than 500 customers across 33 countries are seeing the benefits of Cisco UCS + NetApp. In fact, I”ve blogged about European FlexPod customers including Accenture, Börse Stuttgart, Computacenter, Terremark, Guiness Partnership, Loughborough University, and many more.
This week at Cisco Live London 2012 you’ll have the opportunity to hear directly from several organizations transforming their infrastructures and businesses on FlexPod and talk with variety of partners activity selling and developing solutions built on FlexPod. NetApp is a Platinum sponsor of Cisco Live and I’ll be at NetApp Stand P1 with the rest of the team for the 4th year. Highlights include:
A few months ago, after a my previous blogs discussing cloud computing adoption, I changed subject and authored a short series of articles around the challenges of adopting an architectural-led approach to your IT strategy in general, and data center design in particular. (If you missed them, you can read them here: part 1, part 2, and part 3). The theme of these articles centered on the Winchester House in San Jose, California.
This house was extended by builder after builder, without any architectural blueprint. Consequently, this house had many doors opening into blank walls, abandoned staircases, and other “features” — and it was in construction for year after year, with point additions compounding the problems. I then asserted that this analogy can apply to how IT architectures sometimes evolve -- bit by bit, without a formal blueprint or “grand master” plan, if you will.
Architecture-Led Facebook Poll Results 31 Jan 2012
I finished the series with a poll on our Cisco Data Center Facebook page - thanks to all of you who spotted the poll and took the time to respond. The results were indeed interesting, so I thought I’d share back the results with you and discuss the implications. As the diagram shows, you certainly told us loud and clear what your biggest issue was when it came to adopting an architectural-led approach to your IT strategy and data center design: “We don’t have clear enough business goals for IT” scooped 65% of your votes, way ahead of all other options (!!) -- so let’s discuss now in some more detail.
Our team is excited and ready for a great week at Cisco Live Europe! We’ve been working on demos, speaking sessions, and social events for the conference – and it’s finally here.
There are several conference sessions on intelligent automation and cloud computing. Follow us at @CiscoIA on Twitter where we’ll post updates and reminders about key sessions.
In fact, you may have to choose between some great breakout sessions being held at the same time. Here are a few of the key sessions that feature Intelligent Automation:
Converged Infrastructure and Orchestration with Vblock and Cisco Intelligent Automation -- BRKSPS-2202 (Tuesday, Jan 31, 14:15 pm)
Cloud Automation -- BRKNMS-2659 (Friday, Feb 3, 9:00 am)
Create “Network Containers” in a Multi-Tenant Data Center -- BRKNMS-3999 (Friday, Feb 3, 9:00 am)
Orchestration of UCS via Cisco Process Orchestrator -- BRKDCT-3105 (Friday, Feb 3, 11:00 am)
At Cisco Live this week, you’ll learn how our Unified Management solutions deliver intelligent automation for intelligent infrastructure solutions in a Unified Data Center approach:
When you’re ready to unwind after the big first day, join the data centre team at 18:30 for a meet-up at the W XYZ bar in the Aloft ExCel Hotel next to the conference. Here’s your personal invitation.
And for even more fun, play the Cisco pinball in the World of Solutions! There are prizes for high scores every day, with a pinball challenge on Wednesday at 16:00 – follow @CiscoPinball on Twitter for details.
We look forward to meeting you – enjoy the conference!
The other day I was “making” breakfast for my daughter, as I poured the cereal from the box I noticed that the “toy” was the box. The box had a game board printed on it and you had to cut it out along with the game pieces and the die. The game instructions were printed inside the box, so I ended up with a whole bunch of cardboard snippets and the game. She was so excited to play the game, so we played the game, it was an arduous journey but I stuck it out because that’s the kind of dad I am. When we finished playing, my daughter with her eight years of life experience remarked, “You just don’t get anything good for free anymore!”
Anymore?!! How many reference points could she have in her eight years? Myself, being a little older, I have some more experience with “free”, and sadly I have to agree with her that the free you get today is not as good as the free you used to get, to the point where seeing the word free no longer excites me.
Well I have found some exceptions to that rule. As a developer for a couple decades now I have used many development tools and environments. Along the way either I or my employer had to pay for those tools, but now almost every development product I use is free and not cheesy cereal box game free but full featured with community support and add-ons to extend the capabilities of the product. That’s the kind of free to get excited about.
Why get excited? Well take for instance the Cisco UCS Manager; it has an XML API that is communicated with via XML documents in an HTTP post. Every tool I use to write applications that interface with the UCS Manager is free. That’s great you say, but you still need a UCS Manager and to get that you need a UCS Fabric Interconnect and you need chassis, blades, adaptors, memory, disks, etc. and you start to think things like “the UCS we just ordered isn’t here yet” or “the UCS systems we have are all production and we can’t test our scripts against them.”
Don’t despair because Cisco has released a UCS Emulator for the 1.4 and 2.0 versions of the UCS Manager firmware. The UCS Platform Emulator (UCSPE) has significant capabilities to enable UCS Manager development, things like hardware configuration importing from a live system, exporting of a configuration and the ability to build an emulated UCS system like the one you just ordered. In addition to the emulation capabilities the UCSPE also includes the UCS Object Model documentation, example scripts, Visore (the object browser) and more. Don’t know what the Object Model is or concerned you need more information on the XML API, check out these docs. Read More »
Cisco IASBU has recently launched an update to their In-Memory appliance offerings. Cisco IASBU has certified additional Cisco UCS based systems that now support an entry level SAP BWA appliance with SSDs for storage. In addition, in order to broaden even more coverage, a 12 and 16 blade configuration for SAP BWA that uses NetApp storage. However, what is really driving buzz is the inclusion of the Cisco Intelligent Automation (IA) for In-Memory as part of all the BWA offerings. The same core technology that drives the Cisco IA Cloud orchestration solution. This new automation content supports SAP BWA for Wave 2 customers (non-traditional SAP customers), SAP BusinessObjects (BOBJ) and will serve as a foundation for SAP HANA 1.0 automation content.
On several recent occasions, in discussions with my customers, colleagues and industry peers, the importance of the network, as it relates to Cloud Computing and Data Centers, has been challenged. I am surprised that such a topic is even up for debate ! In my opinion, the underlying network infrastructure of any given Data Center is the architectural foundation for service and application strategy; be it Cloud Computing, Virtual Desktops, Video or even Hosting services.
If we look at a broader scale, no one can argue the complexity and at the same time, the intelligence the modern Internet brings to it’s consumers. How would enterprises and service providers alike, offer converged services like voice, video and data without any network intelligence ? Not to mention, security, application scaling and other managed services. Networks are no longer the traditional packet switching platforms, it’s the heart and soul of intelligence which integrates with other intelligent applications to differentiate the multitude of services that can be enabled over a common medium. As application requirements are increasingly becoming complex, the need for equally smarter transport is critical.
Virtualization is bringing a whole new perspective to this discussion. It’s true you can account for network, compute and storage virtualization within a given solution; virtual switch, virtual machine, virtual firewall, virtual load-balancer, etc.; but how far can we abstract the network ? One can absolutely argue, Cloud Computing is server/compute resource centric, however for most enterprises, when you combine this compute structure with application workload requirements from business, technology and operations perspectives, suddenly the foundation architecture plays a crucial role -- i.e. the network and it’s interconnects.
