2An Introduction to Avaya Fabric Thanks for meeting with us today. We want to use our time today as efficiently as possible. Now, we designed a presentation that is a little bit different than what you may have seen in the past; the presentation is meant to engender an open discussion. Each slide presents one technology solution. The information and diagrams allow us to have a brief or more detailed discussion depending upon your interest level. So, as I present the information, please help me understand your interest level for each technology.Alex WilliamsNetwork Specialist – Western Region
3The simple truthNetworks are largely managed and operated as they were over ten or more years agoBrute force, CLI, scripts, spreadsheets and chasing spanning tree or OSPF gremlins is more often than not, the norm.…why?
4“IT budgeting is both art and science, combining financial management expertise and executive judgment”Gartner, 2012
5But the nagging question is why…. Why are we still doing moves, adds & changes manually?Why do we “live with” unstable networks causing application downtime?Why do we build “overlay” network for critical applications?Why, when OPEX = 80% budget, do we settle for complexity?Why, when labor is the largest line-item expense, do we make it so labor-intensive to plan, configure & maintain a network?And why hasn’t anyone done something about this?Where’s the ‘Executive Judgment’ side of IT budgeting?
6“We need 1,000 more security cameras up by Friday” ThroughputResiliencyLatencyDensity“We need 1,000 more security cameras up by Friday”“Our financial transactions need to be isolated by next week”“We need to transfer a live server instance now!”“We need to have confidence of uninterrupted service”When many of us started out in the networking industry, this is how evaluated solutions. We simply compared fact sheets and drew conclusions. However, today, things have changed. The business has realized the strategic importance of the network. The business can be very vocal and here are the first things out of their mouths… [Next Slide]
7Zeus Kerravala, ZK Research The Bold Truth“It is nearly impossible for today’s networks to run at business speed.”Zeus Kerravala, ZK ResearchWe wait…Minutes for network re- convergenceHours for troubleshootingWeeks for maintenance windowsMonths to roll out significant applicationsIn fact the bold truth is: It is impossible for today’s networks to run at business speed.. [click] [verbally highlight bullets] There needs to be a better way… [Next slide]
81 9 2 8 3 4 7 6 5 Would you be Interested? What if you could… The old network needs to evolve to support today’s business speed92What if you could…83Add HD video surveillance and multicast sites in minutesDeploy services on the flyCreate unlimited separated services with minimal effortDeliver unified access solutions at the edgeEnable Guest, BYOD Access control all at onceIncrease and monitor real-time Network qualityEnable data center continuity within clicksReduce latency and time to service in the data centerDeliver collaboration applications in a stack74We have been building networks the same way since the mid 80’s, and we know that it is not good enough. What if I told you that you could…[go through the nine opportunities]Which of these have the most interest to you? [Explain that you will speak about those solutions first and introduce the others as you go along][Next slide]65Would you be Interested?
9Business Speed Requires Agile Network Control Key Network AdvantagesPhysically Managed Control PlaneAutonomic Control PlaneNetwork-wide VLAN tagging or multiple routing instancesNeed for sequentially dependent legacy protocolsNetwork loopsEdge-only provisioningAll links active/optimizedMulticast scale/performanceEasy and unlimited network segmentionVirtualize devices into one network entityServicePresentation DialogueBefore we get into the solutions, I need to introduce a new and fundamental change in networking. This fundamental change will positively impact almost every other element of your network.In today’s network, services are provisioned against physical devices. Data packets stop at each device on the path, get inspected, and are then forwarded on. This kind of model creates at least three significant constraints that hinder business speed.The 3 C’s of physically managed control plane:Configuration – Relegated to box by box configuration (even in the core – yikes!)Complexity – Have to keep track of all the configurations across all the devices, must maintain multiple routing instances, must have a anti-looping protocol (spanning tree), and the L2 (xSTP), L3 (OSPF), and multicast protocol must all converge in sequence.*Cost – The cost of recovery time and the cost of capacity – With physically managed control planes recovery performance is slow since idle links have to be located and brought into action. Idle links cost money; nobody likes paying for something that is seldom used. Further, multiple protocols means a greater need for CPU power; more protocols, more routing tables, inspection of every packet, and flow balancing all take their toll. [click]There is a better way. Look what happens when physical devices are virtualized into one network entity using a multi-path link-state protocol called, Fabric Connect. [click] Now, network services are abstracted from the physical devices. Services are now provisioned against a network entity, not against physical boxes. The traditional data packet is simply encapsulated at the edge and forwarded to the destination without intermediate inspection. With this in place, your control plane can now be autonomic (borrowing from the concept of the body’s autonomic nervous system, where life support is unconscious, automatic, and involuntary – it just happens). [click]So, the following becomes autonomic with a fabric-based virtualized network. [review bullets]From a technical perspective the following advancements happen in your network. [review bullets][Next slide]*Note - A more exhaustive list of technologies and protocols that Fabric Connect replaces: xSTP, OSPF, Static routes, BGP, PIM-SM, DVRMP etc. (depending upon the need of the network, not all would likely be replaced, but at least three is common.ServiceIP PacketsServices are provisioned to physical devicesServices are provisioned to a single virtual networkThe following become autonomic:IP PacketsConfiguration of the coreNetwork segmentationOptimized recoveryMulticast routingVirtual machine connectivityLoad balancing of flowsIP Packets
10Rome wasn’t built in a day… neither was the “network” Network “Constructs” Stacked ProtocolsInstabilityNumber of control planesMultiple protocols to support:Loop prevention, L2& L3 forwarding, management, virtualization & multicastProtocols run independently & are provisioned and managed separately.This can cause flapping and instability.802.1StabilityEthernet has been built upon “one floor at a time” Adding on when needed
11Fabric Attach – How does it work? Baseline SetupFabric (SPB cloud) is preconfigured with servicesFA Client and FA Server need not have ISIDs, VLANs, port membership, uplink/trunk ports and tagging, etc., configuredFA Client need not have IP and mgmt VLAN configured (i.e., vanilla switch “out of the box”)Zero Touch ProvisioningFA Client auto attaches to the FA Server to obtain mgmt VLAN and IP configEnd points and client connections trigger auto configuration of ISIDs, VLANs, port membership, etc., on FA Client and FA ServerEnd points and clients are automatically attached to authorized servicesFabric Attach is fully integrated into the overall Avaya Networking solutionFabric Attach builds on top of the Fabric Connect architectureIdentity Engines is the centralized policy decision point for BYOD management and NAC
12Secure Guest and BYOD Networking Use Case – Unified User Access Identity EnginesMulti-vendor solutionsManual integrationIndependent security layersWired and wireless accessSecure employee and guest access wired and wirelessAutomatic VLAN / QoS / VSN AssignmentSingle Sign-on for Aura ApplicationsReporting and analytics for compliancePresentation DialogueTo enable a unified access experience for users today, a multi-vendor collection of various technologies are needed. Most common are: Collaboration applications like desktop conferencing, network access control, guest access, and a network security layer. [click] Although, this combination can work for many companies, their seems to be tradeoffs for each combination. [click]Avaya’s unified access is robust and includes collaboration applications and an elegant automated network-wide security architecture. [click] With Fabric Connect as the foundation, both the user experience and the IT burden can be positive. [click] In this example, a user logs into the network. Based upon her device, medium, and corporate relationship, Identity Engines assigned her to a virtual network, in this case the employee mobile zone. Later, a guest [click] selects the corporate SSID and Identity Engines asks for a mobile number or address where their customized password can be sent. The user logs on and gets internet access, but no corporate services.10 Second Case Study claim information: Case study in progress – claim is based off actual user experience at EBC’s.More solution information:One of only 2 vendors offering complete collaboration BYOD solutionTruly vendor-agnostic and unified wired & wireless solutionGranular policies – easy provisioning – guest access in less than 10 secondsManage access of any user, any device and Aura application with single sign-onOptimized and automated end-to-end across Fabric ConnectWLAN 8100 delivers industry leading voice and video scalability over Wi-FiEmployee Mobile ZoneGuest ZoneCompetition’s Independent SolutionsIdentity Engines with Fabric Connect
13Multicast Application Multicast Application Superior Application Performance Use Case – Video Surveillance and Multicast ApplicationsSlow network recoveryStrains CPULow scalabilityComplicated troubleshootingOne seamless protocol network- wideRemoves scalability limitsFaster time to servicePresentation Dialogue:Let’s jump into the first solution. Here is a typical physical view of a network today. [click] And here is the body of protocols that are used at the most fundamental level to run the network. To prevent Layer 2 loops, Spanning Tree must be employed; for Layer 3 Routing, OSPF is run, and for IP Multicast you implement PIM and its host of supporting components. Depending upon your needs you may also use: Static Routes, BGP, DVRMP, MPLS, or OTV.So, let’s look at a use case. [click] Let’s say that you are deploying video surveillance or another application that calls for Multicast support. There are two big problems with the legacy model. First, working with multiple interworking protocols becomes complex for implementation, maintenance, and troubleshooting. Think of deploying, not hundreds of video surveillance cameras, but thousands. Second, when the network re-converges all the protocols must converge in sequence – this costs valuable time. Again, with thousands of cameras and Multicast streams, you are not just dropping VoIP and video sessions, but you are now losing video footage… and likely when you need it the most. [click]Now look what happens when those legacy protocols are replaced with a single next-generation enterprise-wide protocol called, Fabric Connect. [click] It is much simpler and much more powerful. Now cameras and devices are provisioned to the Fabric via the first Switch on the Edge of the network. The Fabric automatically takes care of the rest. [click]One of our Customers, in a recent third party podcast, shared their experience and results using Fabric Connect. They said that they worked very hard with their previous OSPF implementation to get re-convergence times down to a couple of seconds, but they reported that users still noticed it network recovery issues. Now, with Fabric Connect, their inter-site failovers are under 20 milliseconds. Do the math - that is a factor of 100x improvement. Let’s look at another use case. [Next slide]Leeds Metropolitan University Case Study claim information: This claim is from a public podcast (Packet Pushers). Here is the quote: “The original driver for an Avaya Fabric was for L2 extension between Data Centers; however, as we became familiar with the IP capabilities of Fabric Connect we looked at how our routing could be enhanced. With our previous OSPF implementation we got re-convergence times down to a couple of seconds - but users noticed it. With the IS-IS foundation of Avaya’s Fabric we have reduced inter-site failovers to under 20 milliseconds. We can now keep VoIP calls up, mailboxes don’t fall over. Its seamless.” - Phil Taylor, Communications Consultant from Leeds Metropolitan University Based on the quote, here is the math: Math: 2000/20 = 100.Additional Market and Technical Notes:Video surveillance is a key component of the safety and security procedures of many organizations, providing real-time monitoring of the environment, people, and assets and offering historical recording for investigative purposes.The market is poised for rapid growth (24% through 2016) ; recent events have only fueled the growth.Video Surveillance market is transitioning from Digital to IP. Customers want to capitalize on low cost IP video surveillance cameras however most do not have a network that can easily support itIP Video Surveillance needs IP Multicast to function fully and efficiently. For many network engineers, this can mean a lot of headaches. For instance,Main limitations of today’s IP Multicast networks:Multicast stresses switch CPU in larger networks therefore parallel networks are builtLoss of key video and gaps in archived materials due to slow recovery times (Case in point - a nuclear power plant and the network re-convergence was 35 minutes.Limited scale – caps off at 1,000-2,000 camerasBecause so many legacy protocols are employed, the solution is complex for deployment & troubleshootingWhat if Avaya told you, there was a better way? The Avaya Multicast over Fabric Connect solution provides unmatched scalability, flexibility and ease of deployment unlike any competitive offering. Finally - a pill for your IP Multicast headacheThe ability to provide multicast over Fabric Connect is already inherent in the protocol. You simply turn Multicasting on, no configuration or tuning is required.The key advantages to this are:Low CPU usage since there is little protocol overheadLegacy protocols supporting multicast are interdependent and rely on transmitting huge amounts of routing data between protocols that consume large amounts of CPU. So, instead of a big PIM tree getting built out, and a lot of decisions being made, you don’t have to do a reverse path forwarding check to see if you should be sending traffic out on a particular interface because SPB is sorting this all out by relying on the topology calculations of IS-IS where the source and destinations are joining into the Fabric. (In a traditional PIM environment you have to look for the reverse path and they have to search for routes through the routing table until they find the interface that is the reverse path and that is why the CPU hit is so high).Sub second recovery means always-on and no lost dataAvaya Fabric Connect solution converges quickly because there is only one control plane (IS-IS works for Unicast & Multicast)Scalable to tens of thousands of cameras without performance decreaseWith Fabric Connect, bridging and routing take place at the network edge (nodes at the edge convert the IGMP join requests into service membership participation requests. Any receiver can join a Multicast stream efficiently and rapidly) and allows for building of optimized Multicast forwarding (L2 & L3)Single protocol versus complex interdependent protocol overlaysThe widespread use of IP Multicasting belies its complexity. The technologies needed to make Multicasting work over a IP network are complicated, involving protocol overlays that must be kept rigorously in synch with underlying network topologies. With Avaya Fabric Connect, there is one unified protocol – SPB (Shortest Path Bridging)Single command configuration versus complex network-wide configurationTo configure Avaya’s solution you simply type “Enable Multicast” on end-points. Traditional IP Multicast solutions are notoriously complex to configure - PIM must be enabled on all routers globally and also enabled on all core interfaces and local edge interfaces where Multicast is required. Rendezvous points and bootstrap routers have to be configured. With Avaya Fabric Connect, you provision at the edge; there is no need to touch the core.One converged network is more cost-effectiveBest thing about Avaya’s solution is that you get all the benefits discussed without needing to build an overlay network. You can support IP Video surveillance on one converged network. Fabric Connect provides a totally private and closed network for your multicast traffic (L2 and L3 VSN)Camera RecordingCamera RecordingMulticast ApplicationMulticast ApplicationCompetition’s Interdependent legacy protocolsAvaya Fabric Connect
14Competition’s Interdependent legacy protocols Superior Service Separation Use Case – Dynamic and Secure Service ProvisioningMust touch every affected deviceSlower to configureMust configure core of networkDependent upon maintenance windowsEdge-only configurationEliminate need for maintenance windowsMitigate core configuration riskPresentation Dialogue:Most medium sized to large companies have some common network-wide needs once they grow out of a single site. First, they must have high availability across large geographical areas. Second, most require some type of traffic separation, and third, implementation and maintenance of services must be manageable. Industries like transportation, healthcare, and governments need to completely isolate some types of traffic for quality of service, availability, or regulatory compliance. Today, there are really two options to accomplish this: build a separate physical network or use complicated configuration methods (VLAN tagging and VRF’s) with legacy protocols.So, to the use case… let’s say that you are a business that must create completely isolated traffic. You will likely need to meet the Payment Card Industry’s Data Security Standards (PCI DSS) and you may need to create separation between organizations or businesses. [click] You could isolate traffic by configuring every network device, including Core Switches, that would touch the traffic. [click] Or you could only provision at the Edge and let Fabric Connect handle the rest. [click] Because the topology is self-learnt, change and provisioning is self-propagated across the fabric enabling configuration to only be required at the Edge of the network. The network can be fully integrated with automated provisioning tools; scenarios include role-based access management leveraging Avaya’s Identity Engines solution, and server virtualization administration (e.g. VMware’s vCenter) to facilitate the networking requirements in support of compute life cycle management. [click]Since you are no longer configuring the Core, two of the greatest benefits are that you are no longer dependent upon maintenance windows and you mitigate risky configuration errors.One of our customers waiting an average of six weeks for network changes due to maintenance window coordination. Now, from planning to implementation they are done within days. Do the math: 42 days to 2 days. That is a 20x acceleration factor.Fujitsu Case Study Claim Information: Quote: “Previously we needed six weeks for network changes; today we can implement them in days.” Math: 42/2 = 21 ~ 20x Faster Source: Albert Knoll, Fujitsu – Avaya Case Study -For more information on HPAA and PCI Compliance, see Leveraging Stealth Networking to Facilitate PCI-compliance:New ServiceNew ServiceCompetition’s Interdependent legacy protocolsAvaya Fabric Connect
15Superior Virtual Networking Use Case – Multi-Tenancy Extremely complicatedPractically un-scalableError proneStatic modelHighly scalableAgile configurationSimple troubleshootingHighly dynamicPresentation Dialogue:Beyond security, companies have great need to separate services. Think of an airport. Traffic from vendors, the Federal Aviation Administration, individual airlines, baggage, and guests MUST be isolated. Think of a hospital, a university, or government agencies. A fairly common way that this has been done in the past, is to build separate infrastructures for each. Another way to do it is [click x 4 – for each configured legacy pathway] to configure separate paths all the way through the network. As you can see, this would get complicated, wouldn’t scale, and error prone really easily. [click]However, what if provisioning a virtual network was as simple as (in most cases) one command per Switch? What if you didn’t have to worry about anything else? Scaling is no problem, troubleshooting is simplified, and now the vision of robust virtual networking can be met.We have our own great case study for this. In 2013, Avaya was the provider for one of the largest temporary networks in the world, InteropNet for Interop, the leading networking technology show in the world. With only 4 engineers and in three days the network was set up in 1/10 the time from the two prior years with other vendors.InteropNet Case Study Claim Information: “Four system engineers completed the network installation in three days -- 1/10 the resources of previous events.” Wall Street Journal press release: “InteropNet 2013: Unbreakable! Avaya Fabric Connect Delivers on All Fronts” May 15,Competition’s Interdependent legacy protocolsAvaya Fabric Connect
16Access Unification is Changing Integrated Network ServicesPolicyEnabled AccessEnd-to-End FabricsPresentation Dialogue Traditionally, WLAN was deployed as an overlay network in conference rooms and other common work areas, more as a convenience. It was considered a "secondary" network while the wired network was the primary means to connect to the network and access corporate services/resources. However, over the last few years, WLAN has not only become more reliable, but has become very seamless and easy to use. Smart Phones and Tablets have become common place in the work environment where employees use them for work-related purposes and in some cases use them exclusively to accomplish mission-critical tasks in the workplace. And as you’ve probably noticed, these mobile devices do not have Ethernet ports but instead rely purely on WLAN for network connectivity. The next-generation WLAN technologies such as ac bring performance and reliability of wired Ethernet to WLAN. These developments are propelling WLAN from being a secondary network to becoming the primary access mechanism in enterprises. [click]Avaya's end-to-end Fabric technology enables Gigabit WLAN access while eliminating overlay tunnels and the traditional WLAN controller performance and resiliency bottlenecks. [click]As a natural extension of the previous trend we just discussed, network access technologies are becoming "unified." End users expect the same level and quality of access to their authorized resources regardless of whether they are coming in over a wired network or a wireless network. Similarly, network administrators are demanding an integrated way of managing wired and wireless networks and have a common set of policies for security, QoS, etc. Industry analysts have embraced unified access and have stopped covering wired and wireless networks separately in their analyst reports. [click]Avaya is uniquely positioned to deliver unified access because Avaya's Fabric Connect is the only Fabric technology that spans the Data Center, the Campus, and the Edge, and delivers common services all the way to the wireless edge! [click]The third inflection point is BYOD. We are past the stage of saying "No" to BYOD. Most employees today have 1-5 connected devices that they use on a daily basis. BYOD is not only important for employee morale and retention but in a lot of cases, critical to employee productivity. So much so, that several companies are moving away from corporate issued devices and instead providing an allowance to employees to buy their own favorite device. It’s now less about ‘controlling access’ and more a case of ‘managing access’. The key is to BYOD is effective mobile device management (MDM) and security that provides flexibility in access while ensuring that your valuable corporate data assets are protected. [click]Avaya's Identity Engines (IDE) is a central policy decision point for network access. Regardless of whether a device or user is coming over a wired, wireless or a VPN connection, IDE uses a common set of policies for authentication and authorization. Ongoing security extensions and future integration with MDM solutions make it an ideal platform for managing BYOD environments. [Next slide]Unified Access Inflection PointsPrimary AccessUnified AccessBYOD Mission Critical
17Summary: Avaya delivers an outstanding user experience with unified access Avaya FabricConnectState of the art WLAN APs and ManagementAdvanced RF ManagementApplication QoS EnforcementGuest and BYODUnified AccessWi-Fi ChallengeAvaya SolutionAPPSUnpredictable volume of application useUnknown and uncontrolled recreational app useNext generation Application Recognition and Control at the network edgeEnd-to-end proactive quality of experience monitoringUSERS &DEVICESUnpredictable types of BYOD devicesDifferent users with multiple devicesDevice fingerprinting, roaming assist, Bonjour support & secure network and BYOD accessSPECTRUM& NETWORKCAPACITYCongestion in crowded 2.4GHz spectrumSticky clientsOverlay tunnels and controller bottlenecksDisparate wired and Wi-FiFlexible deployment options for spectrum optimization2-tier architecture + Avaya Fabric (when available) + Unified AccessCOMPLEXITYChanging requirements from 2.4GHz to 5GHzComplex planning, deployment & supportSoftware upgradable (11n to 11ac); Ease ofEase of deployment with Cloud offering and Avaya Fabric to the wireless edgeAutomated edge
18Scale of the Olympic Games 6K803 Billion TV ViewersMillions of SpectatorsThousands of OfficialsOlympic AthletesOlympic TeamsTo give you an idea of the scale of an Olympic GamesWe have to meet the diverse technological needs of a demanding audience – the Olympic Family – about 40,0000 accredited athletes, coaches, volunteers, officials, etc.including:80 Olympic teams - made up of over 6 thousand of athletesThere are thousands of officials and scorekeepers14,000 members of the press25,000 volunteers3 billion TV viewersAnd billions more onlineWe expect them to bring about three devices apiece to the event. So that’s going to be over 150,000 tablets, laptops and smartphones that we’re going to have to register when they show up on the first day, identify who they are, secure them and give them access to the right services.Our technology is supporting 200,000 various scoring metrics for competitions.In these type of events, there are hundreds of opportunities to us achieve success.Like the athletes of each game, you “go for the gold” in a highly competitive and aggressive arena…14K25KReporters and MediaVolunteersBillions on the Internet
19The Scale of an Olympic Games Network Interconnecting11 Competition Venues2 Data Centers2 Technical Operations Centers3 Olympic Villages2 Media Centers2 Celebration CentersOlympic Scale Network54 Terabit capable backbone2,000 Ethernet switches50,000 Ethernet ports2,500 Wireless Access Points36 HD Video Channels1,500 IPTV screens6,500 VoIP phonesInterconnecting11 Competition Venues2 Data Centers2 Technical Operations Centers3 Olympic Villages2 Media Centers2 Celebration CentersOlympic Scale Network54 Terabit capable backbone2,000 Ethernet switches50,000 Ethernet ports2,500 Wireless Access Points36 HD Video Channels1,500 IPTV screens6,500 VoIP phones19Similar to running 3 Super Bowls for 17 days straight!
20Get 360o visibility into the Wi-Fi network Next-gen management in action at Sochi Winter Olympic Games
21Seven Secure Virtualized Networks Over a Converged Core CelebrationCentersOlympic VillagesIPTV ServiceGames NetworkUnified Communications ServiceOrganizing Committee ServiceWi-Fi ServicesMedia Transparent LAN ServicesSecure ManagementCompetition venuesThe network we built for the Games consists of multiple distinct networks and applications – which all need to be isolated and separate from each other. The ability for the Avaya Fabric Connect technology to support multiple private virtualized network over a common infrastructure without any additional equipment or protocols was a huge selling feature of our technology. And something that we are uniquely suited for.IPTV service – distributes 36 HD channels of live Olympic Coverage to each of the Olympic venues. It ensures that athletes at the Olympic village can stay connected to all the competition events and medal ceremonies. It is also useful for coaches who may not have the proper accreditation requirements to get into the specific venue to watch and analyze their athletes performances.The Games Network – This network carries, distributes and supports approximately 200,000 critical distance, timing and scoring metrics for each of the competitions. It also support specific Olympic applications that are leveraged for each Games event such as Information Systems for commentators.The Unified Communications Service - Provides communications services at each of the venues for athletes to leverage to call home, journalists to call back to headquarters, volunteers to leverage at help desks and for officials to use at the competition venues to call in a request for a video replay.The Organizing Committee Network – This is the network that supports the individuals who put on the Games.Wi-Fi Services – This network provides free Wi-Fi services for all of the Olympic family and media from each of the venues.Media Transparent LAN Services – When large media firms such as Routers prepare to send their journalists to an event like the Games – they pre-order transparent LAN services which give them a dedicated connection from the main media center at the Games to whatever venues they want to connect to.The last separate environment is a network that is dedicated just for control or management traffic. This was deemed necessary due to the mission critical nature of this environment.Data Center(primary)Data Center(secondary)MediaCenters
22High-Performance Data Center Up to 4x Better Latency and 25x Better Time-to-Service Multi-Tier Legacy Data Center ArchitecturePresentation DialogueApplications are becoming more complex as they require multiple interactions between different server components (web, authentication, application, database), and as they are hosted on multiple virtual machines and physical servers. This is what is driving the dramatic rise in horizontal or east-west traffic volumes, and every one of these interactions is subject to network latency (the multiplier effect). [click] Networks that remain architected north-south will suffer from two related issues; first, the (relatively low capacity) uplinks between tiers – between ToR and Aggregation, and between Aggregation and Core – will be overloaded with local, east-west traffic, and second, the cumulative latency of traversing these tiers and devices will dramatically impact overall application responsiveness. [click]The basis for Avaya’s high-performance Data Center is our innovative Distributed Top-of-Rack solution. Distributed Top-of-Rack (or DToR) delivers a radical solution to the challenge of maintaining application performance in the Data Center.Avaya’s Distributed Top-of-Rack has been specifically designed to complement distributed applications. We leverage dedicated, ultra high-speed interfaces to interconnect multiple ToR Switches that empowers east-west traffic to bypass the north-south traffic jam and go directly to its destination, along the shortest, fastest, and most efficient path. Distributed ToR delivers, on average, delivers a 4x advantage in terms of reduced latency, and it also provides significant cost savings by minimizing or even eliminating very costly Aggregation and Core links and equipment. DToR is fully optimized for modern application design, and because it’s part of the VENA framework is also seamlessly integrates with our Fabric Connect technology to streamline service provisioning, reduce errors, and dramatically improve time-to-service.Technical NotesDToR - scalability up to 200 units and an incredible112TbpsClaims Information"Up to 25x faster configuration"Miercom Report: https://avaya.my.salesforce.com/sfc/servlet.shepherd/document/download/ z5iiAAA?doc_type=Third-Party%20Report"Up to 4x better latency in the data center"Lippis Active-Active Ethernet Fabric Test report. Average latency of 7usec, as opposed to 30usec+ for 3-tier network (“The average latency across the Avaya Distributed Top of Rack solution was 4x faster than the traditional 3-Tier network” - Nick Lippis, Active-Active Cloud Network Fabric Test, April 2013)Distributed ToR / Fabric ConnectData Center ArchitectureBuilt for north-south trafficAdditional tiers brings complexity and costStatic modelHighly scalableAgile configurationSimple troubleshootingOptimized for east-west traffic
23Network-wide Proactive QoS Monitoring Avaya-on-Avaya Access Solution Avaya-on-Avaya AdvantageExpensive with additional overlayReactive using historical miningIncomplete with no end-to-end visibilityProactive monitoring‘Agents everywhere’Graphical reportingPresentation Dialogue:Many organizations struggle to monitor real-time communications on a real-time basis. [click] When session quality is poor, it sets off a string of events that usually seems to end up in IT’s lap. [click] Third party solutions can be procured, but are hard to justify as an overlay. They typically use historical data and struggle to address all end-points. [click]Avaya’s Service Level Agreement Monitor (SLAMon) is a proactive tool that can monitor customer networks to detect problems before applications are affected. [click] For instance, tiny synthetic packets are sent into the network. [click] Data traffic is analyzed and reported to understand how a session would react if it were going. The advantage the Avaya has is that these quality agents are built into our communication devices and Switches. Think of video calls were there is no physical device. The Switch can serve as the proxy for the video client and all collaboration sessions can be monitored.Additional Solution InformationExpensive vs. lightweight (and cost-effective): Avaya’s approach relies on embedded software agents in its IP Phones and Switches, avoiding the need to dedicated hardware probes or overlay monitoring equipment typically required by competitor’s solution. Agents are embedded today in Avaya 96xx Phones, Avaya ERS 3500, 4x00 and 5x00 Edge Switches, as well as Avaya ERS 8800, VSP 7000 and 9000 Switches…with plans for Avaya G430/450 and potentially other devices in the future.Proactive vs Reactive Monitoring: Most vendors’ solutions are reactive - you know about a bad voice/video experience after it happens. Avaya’s solution generates synthetic traffic between specific end-points that can get proactive results about how your network will treat various traffic types under congestion.End-to-end QoS: Avaya’s solution monitors whether all intermediate nodes in the network (across administrative domains, LAN, Service Provider WAN links, etc.) honor the QoS settings or not. It can monitor QoS on a hop-by-hop basis. Competitive solutions lack this ability to build a true end-to-end view of QoS treatment across multiple hops.Poor session qualitySynthetic tests sentCompetition’s Reactive NetworksAvaya SLA MontmCall to IT and troubleshooting beginsTrouble segments identified BEFORE call begins
24Avaya has a comprehensive wired and wireless portfolio Network-WideInfrastructureCollaboration PodFabric ConnectVSP 9010ERS 8000VSP 7000WLAN 9100VSP 4000VSP 8284ERS 4000/5000ERS 3000Unified Access
25Can we share more? Presentation Dialogue Again, thanks for your time today. I have appreciated our discussion. I would like to ask if our discussion has prompted a desire to meet again and go deeper into any of the technologies that we have discussed today.
26Questions?Thanks for meeting with us today. We want to use our time today as efficiently as possible. Now, we designed a presentation that is a little bit different than what you may have seen in the past; the presentation is meant to engender an open discussion. Each slide presents one technology solution. The information and diagrams allow us to have a brief or more detailed discussion depending upon your interest level. So, as I present the information, please help me understand your interest level for each technology.Thank You
28State of the art Wi-Fi Management Avaya WLAN 9100 Orchestration System Network Management and AnalyticsFlexible deployment optionsOn-premise software solutionCloud managementKey benefitsZero-touch deploymentHighly scalable to 1000’s of unitsFlexible toolsets for remote management and troubleshooting
29Next-gen wireless with predictable performance State of the art WLAN APs State of the art WLAN APs that address a broad range of use casesWAP 9122WAP912391329133WAO9122WAP 9162/9173802.11n2x2 AP3x3 AP*802.11ac3x3 AP2x2/3x32 Radios4 RadiosNext-gen wireless with predictable performance802.11acSimplified RF planningApplication QoS controlBYOD/Guest access2-Tier architecture with distributed controllerSwap-less upgradesCloud-based managementEnd to End Unified AccessFlexible deploymentLow TCOBonjour supportCommon Advanced Feature SetIntegrated ControllerApplication ControlZero Touch ProvisioningOn-Premise or Cloud Management
30A New Class of Access Point: High End Features in a Low Cost AP State of the art WLAN APs Comprehensive offering that address a broad range of use casesA New Class of Access Point: High End Features in a Low Cost APINTEGRATED CONTROLLERNo single point of failure or performance bottleneckAPPLICATION CONTROLFull DPI at the network edge – an industry firstSPECTRUM OPTIMIZEDDual 5GHz AP for future proof migrationEnhanced high density support in 2.4GHzWLAN 9122WLAN 9123WLAN 9132WLAN 9133WLAN 9162WLAN 9173802.11n2x2 AP3x3 AP*802.11ac3x3 AP2x2 MultiRadio AP3x3MultiRadio APFixed ChassisModular Chassis2 Radios2 or 4 RadiosNA4 Total SlotsNext-gen wireless with predictable performance802.11acSimplified RF planningApplication QoS controlBYOD/Guest access2-Tier architecture with distributed controllerSwap-less upgradesCloud-based managementEnd to End Unified AccessFlexible deploymentLow TCOBonjour supportCommon Advanced Feature SetIntegrated ControllerApplication ControlZero Touch ProvisioningOn-Premise or Cloud Management
31Advanced RF Management for “wired-like” performance Auto Channel & CellMaximize use of RF resources to improve the user experienceAutomatic radio controlLoad balancing across radiosRoaming assistance for sticky clientsBonjour optimizationHoneyPotIncreased availability, resiliency and application performance802.11ac, MIMO, etc.ALB (Auto-Load Balancing)Roaming Assist
32Avaya Virtual Services Platform 8284XSQ Uncomplicated, compact format & high- performance platformNon-blocking, wire-speed, & latency-optimizedMature, feature-rich, & robust architectureVirtualize just the Core, or go to Fabric Connect for end-to-end virtualizationUnique combination of port speeds/density & features/functionalityHighlightsSimple to deploy & operateFuture-proof performanceProven & feature-rich OSVersatile virtualization capabilitiesReduces TCO & drives faster ROI
33What’s Special..?Ushering in a new era of simplification & efficiencyThe VSP 8284XSQ brings Modular-class performance and functionality to the mid-size Core, with fixed-format levels of price, simplicity, and operational efficiencyCosts less to acquire, dramatically lower Price/PortHigh-density, 80 x 10 Gigabit, plus 40 Gigabit, all in 2RU..!Unique packaging of the Industry’s most advanced chipsetEasier to install and operationalizeDelivers high-end, feature-rich functionalityVersatile conventional or Fabric-enabled networkingMore cost-effective to support and maintainSet-and-forget solution for resource-constrained IT departments
34High-Availability Dual Power Supplies What’s New..?Rear of the new VSP 8284XSQ platformDimensions: 440mm wide x 500mm deep x 88.9mm tall (2U)Weight: kg ( lbs)High-Availability Dual Power SuppliesHigh-AvailabilityQuad Fan Trays
35What are the Numbers..? Specifications & scale Switch Fabric: 2.56Tbps Frame forwarding: 1,428MppsNominal Latency: <480nsecSwitch Cluster/LAG: up to 96 Groups of up to 8 LinksVLANs: up to 4kSpanning Tree Groups: 64MAC Address: up to 224kARP Entries: 32kJumbo Frames: up to 9,600 BytesIP Interfaces: up to 256IPv4 Routes: 16,000 (128k*)OSPF Areas: up to 80OSPF Neighbors: up to 500ECMP Paths: up to 8VRRP Instances: up to 512VRFs: up to 24IS-IS Neighbors: 64SPB Nodes: 500L2 VSN IDs: 4kSPB IP Shortcut Routes: 16k* Hardware-ready for future field-configurable options leveraging the UFT capability
36What’s the Positioning..? Versatile capabilities make for a flexible offeringIdeal positioning for the VSP 8284XSQ:Core Switch (typically deployed as a high-availability pair) for small- to-medium Campus networksInterconnecting multiple 10GbE-attached Wiring ClosetsDistribution/Aggregation Switch40GbE downlinks and 10GbE uplinksDeployable in a number of Enterprise networking environments:Conventional, multi-tier IP Routed Campus networksFabric-based Campus networksAdditionally, the high-performance architecture means that it can be positioned as:Data Center End-of-Row SwitchTier 3/4 Service Provider & Smaller Hosting Provider networksVerticals: All – very broad appeal, everywhere that values simplicity
37What’s New..?Delivering ‘Networking 2.0’, and do it todayMid-sized Core solution that focuses on delivering revolutionary agility, simplicity, and operational efficiencyDramatically reduces the operational burden and delivers an accelerated return on investmentSeries highlights:tightly integrated hardware and software architectureversatile virtualization capabilitiesfuture-ready performance
38Automating network access has direct impact on reducing cost of change IP PhoneVisitor or Business PartnerPersonal MachineCorporate DesktopNetwork PrinterNetwork DeviceWireless Access PointSurveillance CameraFax MachineMedical DeviceLocal Server/AppGuests & Guest DevicesEnterprise NetworkEach access port is not assigned until a user/device attempts access.Once authenticated & authorized, user/device is granted appropriate access level.MAC address lookup:Ignition Server local storeManual inputWildcards (e.g. Avaya IP Phones 00:04:0d* and Cisco IP Phones 00:15:62*)Import CSV file with list of MAC address and other device attributesAccess Portal auto-populateNOTE: Planned to extended lookup to allow interfacing a LDAP data store
39Real life proof point … The largest temporary network in the world! Interop Cisco sent 20 network engineers to hot stageInterop Huawei sent 25 network engineers to hot stageInterop Avaya sent 3 network engineers to hot stage“Four system engineers completed the network installation in three days - 1/10 the resources of previous events. Then, for the next five days, the show tested the mettle of Shortest Path Bridging …Avaya Fabric Connect met the challenge without a single incident related to the Avaya solution. Layer 2, Layer 3, Unicast and Multicast services ran flawlessly over Avaya Fabric Connect the entire event. Mission accomplished.”“This year's InteropNet demonstrated that, ideologically motivated debating points aside, SPB just works.”Avaya - Confidential 2014
4080 Sochi - The Scale of an Olympic Games 6K 25K 14K 3 Billion TV ViewersMillions of Live SpectatorsThousands of Officials and ScorekeepersOlympic Athletes6KOlympic Teams80To give you an idea of the scale of an Olympic GamesWe have to meet the diverse technological needs of a demanding audience, including:80 Olympic teams - made up of over 6 thousand of athletesThere are thousands of officials and scorekeepers14,000 members of the press25,000 volunteers3 billion TV viewersAnd billions more onlineOur technology is supporting 200,000 various scoring metrics for competitionsReporters and Media14KVolunteers25KBillions of Internet Customers
41Changing Social Media Landscape The online experience of the Olympics is dynamic and changing incredibly fast. Vancouver was called the “the first social media Olympics.” That year, Twitter and Facebook were still quite new and the iPhone 3G was an innovative technologyWired access outnumbered wireless access 4:1. We are planning for the reverse in 2014 at Sochi and are planning capacity in our Wi-Fi capabilities to support 120,000 wireless devices for the Olympic Family only. Fans will access the Internet through wireless 3G services.We also know that athletes, journalists, and other Olympic family members will use social media to share the stories that matter to them—the great victories, crushing defeats and celebrations.
42First “Fabric Enabled” Games DATA CENTERDynamic Data CenterGeographically redundant virtualized Data CentersDistributed Top of Rack for low latency v-motionAvaya Aura for communicationsVENUESIntelligent EdgeFabric Connect extended into each venueAny Port Any ServiceScalable Wi-FiSecure network accessNETWORK CORENetwork VirtualizationAvaya Fabric Connect (based onIEEE Shortest Path Bridging)Builds a cloud based infrastructureHigh capacity / scaling3rd party validated 20 ms recovery ratesMountain Media CenterSwitch ClusteringAvayaVENAPrimary Data CenterSwitch ClusteringAvayaVENAVirtual Services FabricAvayaVENAKrasnaya PolyanaSecondary Data CenterOur design will serve the needs of every audience and meet the highest standards. We built a private network cloud architecture with intelligent edge devices that plug into a robust core network that connects the data centers with each of the venuesAvaya’s intelligent edge technology will provide distributed routing services, Wireless LAN at the venue and Olympic Village, any port, any service access, so no matter where a user connects from – they will get access to their own secure network resources.The network core will be designed with redundancy, deployment flexibility, and Ease of Use as primary considerations. Avaya Fabric Connect will provide the ability to efficiently provision network services with only a few mouse clicks. This will give the Olympic Committee total control, while requiring minimal management resources. It also provides the resilient foundation necessary for such a high profile event with 3rd party validated 20 millisecond recovery rates.There will be dual Data Centers and dual Technical Operations which are geographically redundant and virtualized. To ensure fast performance for latency critical applications we will be leveraging our Distributed Top of Rack functionality which builds a high speed virtual backplane between server racks to deliver low latency switching for east west traffic in the Data Center.Also in the Data Center will be Avaya Aura with Contact Center Elite. This will be used for UC services from each of the venues.. Including competition venues. The Contact Center will be for the IT help desk.
43Avaya Solutions Customers 97%95%90%World’s Top . . .9 out of 10 Airlines9 out of 10 Financial Institutions8 out of 10 Auto Manufactures8 out of 10 Luxury Hotels
45Next-gen wireless with predictable performance State of the art WLAN APs State of the art WLAN APs that address a broad range of use casesWAP 9122WAP912391329133WAO9122WAP 9162/9173802.11n2x2 AP3x3 AP*802.11ac3x3 AP2x2/3x32 Radios4 RadiosNext-gen wireless with predictable performance802.11acSimplified RF planningApplication QoS controlBYOD/Guest access2-Tier architecture with distributed controllerSwap-less upgradesCloud-based managementEnd to End Unified AccessFlexible deploymentLow TCOBonjour supportCommon Advanced Feature SetIntegrated ControllerApplication ControlZero Touch ProvisioningOn-Premise or Cloud Management
46Avaya Virtual Services Platform 8284XSQ Uncomplicated, compact format & high- performance platformNon-blocking, wire-speed, & latency-optimizedMature, feature-rich, & robust architectureVirtualize just the Core, or go to Fabric Connect for end-to-end virtualizationUnique combination of port speeds/density & features/functionalityHighlightsSimple to deploy & operateFuture-proof performanceProven & feature-rich OSVersatile virtualization capabilitiesReduces TCO & drives faster ROI
47What’s Special..?Ushering in a new era of simplification & efficiencyThe VSP 8284XSQ brings Modular-class performance and functionality to the mid-size Core, with fixed-format levels of price, simplicity, and operational efficiencyCosts less to acquire, dramatically lower Price/PortHigh-density, 80 x 10 Gigabit, plus 40 Gigabit, all in 2RU..!Unique packaging of the Industry’s most advanced chipsetEasier to install and operationalizeDelivers high-end, feature-rich functionalityVersatile conventional or Fabric-enabled networkingMore cost-effective to support and maintainSet-and-forget solution for resource-constrained IT departments
48Avaya Virtual Services Platform 4450GSX VSP 4000 is a series of fabric- based multi-service / multi- tenant Edge SwitchesUsed in Campus aggregation network or connecting remote sites over L2 Metro Ethernet/WAN servicesNew VSP 4450GSX-PWR+ delivers a fiber optimized version (complementing copper form-factor products)Plays a key role in video surveillance, video distribution or digital signage.HighlightsSimple to deploy & operateProven, carrier-grade OSBest-in-class virtualization capabilities; full multi-service featuresReduces TCO & drives faster ROISmall distribution – fiber and risersGig rings in the metroGig hand off in the risersOther parts of the globe where fiber is more prevalent.VSP OS – WAN edge or Campus aggregation – more fiber hand-off.Basement of the building; Gig to each floor. Built in copper ports – risers to aggregate fiber. Always copper connections coming into this box.2x10 gig are MacSec – big in EMEA. Aggregate the risers.. But when uplinks leave the building – MacSec capabilities.Metro Ethernet: Tier 3 / 4 SP’s - Drop into building run up the risersCampus aggregation where there is alot of fiber.MacSec encryption for 10 Gig ports. Once you leave the building – need encryption. In the building – its ok don’t need encryption.12 x 10/100/1000 Copperwith PoE+2 x 10 Gigabit SFP+36 x Gigabit SFP
49Avaya Virtual Services Platform 7000 Series Overview & HighlightsVSP 7024XT Switch24-ports of 10GBASE-T RJ45 SocketsFit-for-Purpose for TodayVersatile support for 1/10 Gigabit Ethernet, Copper or TransceiverDistributed Top-of-Rack delivers Industry’s fastest virtual backplaneFabric networking delivered directly to the ServerMedia Dependent Adaptor flexibilityLossless hardware & software architectureFront-back or back-to-front coolingSmall form-factor & energy-efficientFuture-Ready for TomorrowSeamless integration of 40 and (future) 100 GigabitData Centre Bridging-ready to Storage Networking convergenceLightning-fast performanceFlexible connectivity optionsDelivering mass 1/10 Gigabit, plus 40 GigabitFuture-ready for 100 GigabitHighlightsOverview & BenefitsThe Avaya Virtual Services Platform 7000 is ideally suited to delivering today’s pressing need for flexible, high-speed Ethernet connectivity in the high-performance Data Center Top-of-Rack (ToR) role; additionally it provides a cost-effective 10 Gigabit Ethernet fan-out capability for existing Core Switch deployments and it also has application in the Campus distribution layer, delivering flexible connectivity and consolidation options. Featuring a hardened physical architecture of dual, hot-swappable AC or DC power supplies and fan trays, the Virtual Services Platform 7000 (VSP 7000) is an important addition to a network manager’s toolkit for creating always-on high-performance solutions.In addition to the versatility afforded by the integrated MDA, the VSP 7000 boast a host for strategic innovations that ensure it will continue to deliver cutting-edge performance and services throughout an extended life cycle. Chiefly amongst the future-ready features is the advanced fifth generation application-specific integrated circuit (ASIC) chipset that has been optimized for sophisticated network operations.
50ERS 3549GTS Introducing 48-port models Two new models that extend SME & mid-market positioning of ERS family48-port 10/100/1000 PoE+ and non-PoE1x SFP+ (10 GB) uplinkResilient stacking of up to 8 units (up to 384 user / 392 total ports)Easy to installIP Office Quick start (run IP Office - autoscript)IP Phone plug and playLifetime WarrantyCompetitively priced ($3,095 APL; PWR+ $3,660 APL)ERS 3549GTSERS 3549GTS-PWR+
51ERS 3500 Release 5.2 New Software Features Avaya Energy SaverAutomatic power dimming of portsReduces power consumption by up to 25%SLA Mon™ Phase 2 agentCompatible with Avaya Diagnostic Server R2 (April GA)Also supports end-to-end QoS tests via CLI/EDMSimple Loop Port Protection Guard (SLPP Guard)Serviceability enhancementsAuto-restoration and back-up file configuration managementUnified Password SecurityStatic LACP Key to Trunk ID bindingFlash History enhancements