Issue1 Final

download Issue1 Final

of 14

Transcript of Issue1 Final

  • 8/6/2019 Issue1 Final

    1/14

    Creating a Fit forPurpose NetworkArchitecture

    Featuring research from

    Issue 1

  • 8/6/2019 Issue1 Final

    2/14

    2

    Avayas Data Solutions is trusted by many o the most prestigiouscompanies in the world. This trust is made evident by the act that

    Avaya is the only company to ship both 100 million voice lines and

    100 million data ports, has the second largest core switch installed

    base at 2.7 million ports and has the most advanced line o stacka

    switches in the industry having sold over 22 million ports in the la

    our years.

    With Avayas data solutions, we can help enterprises more easily

    adapt to business needs by delivering immediate and long term

    value. With solutions that provide seven times more resiliency, 20

    times better perormance, 40 percent more energy eciency and

    approximately 50 percent better total cost o ownership, customers

    can count on Avaya to improve the value they derive rom theirtechnology investments.

    Steve Bandrowczak, Vice President and General Manager,

    Avaya Data Solutions

    Avayas vision or the enterprisecalls or a new level o

    synergy between people,

    the collaborative real-time

    applications they use, and the

    underlying, enabling network.

    A key building block or this

    vision is the oundational

    network. As real-time

    communications continue

    the evolution to IP, the data

    network becomes completely

    integrated into the delivery

    o communications-enabledbusiness services and mission

    critical business applications.

    Avayas Fit or Purpose strategy is resonating in the market.

    Customers are telling us they want unique solutions that are tailored

    and best-designed to support their specic communications

    requirements. Todays CIOs are ocused on saving money, driving

    new revenue streams, increasing the loyalty o their customers

    and creating competitive dierentiation. Our go-to-market strategy

    is ocused on addressing these needs with some o the most

    competitive products in the industry.

    Introduction

    Creating a Fit or Purpose Network Architecture is published by Avaya. Editorial supplied by Avaya is independent o Gartner analysis. All Gartner research is 2010 by Gartner, Inc. and/or its Afliates. All rights

    reserved. All Gartner materials are used with Gartners permission and in no way does the use or publication o Gartner research indicate Gartners endorsement o Avayas products and/or strategies. Reproductio

    and distribution o this publication in any orm without prior written permission is orbidden. The inormation contained herein has been obtained rom sources believed to be reliable. Gartner disclaims all warrantie

    as to the accuracy, completeness or adequacy o such inormation. Gartner shall have no liability or errors, omissions or inadequacies in the inormation contained herein or or interpretations thereo. The reader

    assumes sole responsibility or the selection o these materials to achieve its intended results. The opinions expressed herein are subject to change without notice.

  • 8/6/2019 Issue1 Final

    3/14

    3

    other changes are also transorming theenvironment in which enterprise networks

    operate:

    IT departments are under greater pressure

    to account or ROI in their expendituresa

    trend that accelerated dramatically with

    the economic downturn in 2008.

    There is a greater recognition of the

    economic cost and environmental impact

    o IT energy consumption. IT solutions that

    are energy ecient are here to stay.

    A huge generational shift is taking place

    in todays workorce: as the Baby Boomers

    retire, their successors are bringing new

    expectations concerning IT availability,

    perormance and fexibility. This will usher

    in a quantum shit rom the traditional

    network to one driven by and built or real-

    time, multimedia, always-on, omnipresent,

    multi-platorm communications.

    Individually, any one o these trends would

    have a major impact on network strategy.

    Collectively, they represent a signicant shitin todays enterprise network operating

    environment.

    2. THE CHALLENGE FACING GENERALPURPOSE ENTERPRISE NETWORKS

    Because todays general purpose network

    architectures were not designed or the

    demands o Unied Communications, the

    traditional tactics or enhancing network

    perormanceadding more hardware,

    upgrading to aster processors, increasing

    bandwidthcan actually intensiy problems,

    such as server creep and congurationdrit.

    The net eect o these and other issues is

    higher network TCO (Total Cost o Ownership).

    Taking an inherently inecient, over-

    engineered approach to network challenges

    absorbs more resources, delivering less and

    INTRODUCTION

    Todays enterprise network needs to be a

    highly ocused resource capable o providing

    a mobile and diverse workorce with ecient,

    eective access to the tools and applications

    that drive productivity. Because traditional,

    general purpose network architectures were

    not designed to cost-eectively address this

    new world o Unied Communications, new

    network strategies are needed. Avaya is

    meeting the challenge with Fit or Purpose

    inrastructure capabilities that combine

    perormance on the most demanding

    real-time applications with compellingeconomics.

    1. THE NEW ENVIRONMENT FORENTERPRISE NETWORKS

    A sea change is taking place in enterprise

    networks. The growing use o the network

    or real-time communications and

    collaborationUnied Communications

    requires a dramatic change in network

    architecture.

    The data output by digital devices, such as

    PCs, servers, mobile phones, scanners,supercomputers, and cameras, is growing

    exponentially. Based on Moores Law,

    we can expect the digital universe to

    double every eighteen months, i not

    aster, placing a heavy strain on real-time

    networks.1

    A huge part of the increase in

    network trac is the result o real-

    time applicationsVoIP, personal

    videoconerencing, WiFithat are critical

    to enterprise perormance.

    All of thisdata, applications, presence

    capabilities, etc.must be made

    available to an ever growing mobile

    workorce.

    This transition to Unied Communications

    is not taking place in a vacuum. A host o

    less perormance while incurring more andmore expense. Enterprises that continue to

    rely on these general purpose architectures

    will nd themselves at a competitive

    disadvantage to enterprises that have made

    the shit to what Avaya calls Fit or Purpose

    networksnetworks designed or the

    realities o todays distributed, collaborative

    enterprise.

    3. THE FIT FOR PURPOSE ENTERPRISENETWORK

    A Fit or Purpose solution is one that

    identies the most critical perormancerequirements and delivers the best

    possible solution to address the specic

    individual needs o todays distributed,

    collaborative workorce, meeting the

    challenge o delivering real-time access

    to communications tools, inormation and

    applications.

    A Fit or Purpose network inrastructure vs.

    a traditional general purpose inrastructure

    is like the dierence between a line o

    perormance racing cars and a line o

    sporty, economical sub-compacts designedor everyday consumers. Both, or example,

    have an engine, wheels, brakes and a

    steering mechanism. Both are designed

    with aerodynamic eciency in mind.

    However, the racing car is designed above

    all to achieve speed. Everything about it,

    not just the perormance o the engine but

    the ease with which aulty components

    can be swapped out or reueling take

    place, is purpose built to save precious

    seconds in a competitive environment. The

    sporty subcompact, on the other hand, is

    designed to optimize its appeal to a marketsegmentspeed is just one actor (and,

    given speed limits, not a major one.) A

    wide range o vendor-driven and market-

    driven actors infuence the design o the

    sporty compact, not the least o which is to

    repurpose other o-the-shel components

    rom the manuacturers product line.

    Creating a Fit for Purpose NetworkArchitecture

    1http://www.intel.com/technology/mooreslaw/

  • 8/6/2019 Issue1 Final

    4/14

    4

    In the case o todays enterprise networks,most vendors, in eect, take the sporty

    compact approach: they start out with

    a set o basic building blocks and meet

    specic market needs (e.g., real time

    communications) by layering on more

    architectures and protocols. What results

    are solutions that sacrice simplicity and

    fexibility and, ultimately, TCO.

    In contrast, Avaya has identied what it

    sees as the most critical perormance

    requirements or todays networks

    resiliency, eciency and scalabilityandmade solving those specic challenges the

    driving orce behind its portolio o Fit or

    Purpose enterprise network solutions.

    4. COMPONENTS OF A FIT FOR PURPOSENETWORK ARCHITECTURE

    Because there is no single solution to the

    challenge o network resiliency, eciency

    and scalability, Avaya addresses each

    characteristic uniquely.

    Here is a quick look at some o the keyarchitectural elements o the Avaya Fit or

    Purpose network inrastructure:

    Bringing Resiliency to a New Level

    Avayas Fit or Purpose approach to resiliency

    is based on its Switch Clustering technology

    and also by more eectively managing

    routine maintenance operations as well

    as unplanned outages. These and other

    capabilities enable Avaya to deliver up to

    seven times better resiliency than the market

    share leader, providing uncompromising

    end-to-end availability or real-timeapplications such as VoIP.2

    Switch clustering: This technology

    and architecture, pioneered by Avaya,

    essentially virtualizes the network core.

    Far superior to the one-size-ts-all,

    spanning tree approach, switch clustering

    utilizes all network resources regardless

    o scale and provides sub-second ailover.

    Any disruptions are virtually unnoticeable

    to the user sessions do not ail andapplications do not hang.

    Managing ailures: I a ailure occurs, a

    ailed unit can be disconnected without

    any loss o overall application availabil

    The Avaya Automatic Unit Replacemen

    (AUR) process manages any necessary

    downloads and then brings the switch

    back online; all o this without any need

    or an engineer to provide conguration

    management.

    Eicient Network PerormanceA good business is run just this side o lea

    and mean; a Fit or Purpose network mus

    be designed with the same sensibility. Ava

    solutions are able to deliver up to 20 times

    better perormance with one-third less

    equipment through an architecture design

    to maximize all resources and investment

    eliminating idle or underutilized links,

    equipment and even entire architectural

    tiers.3

    Operating and Technology Ocer at the school, Avaya was

    chosen because when the decision had to be made no

    other network could do what we needed.

    At Coppin State University in Baltimore, the schools

    cutting edge IT program goal which is occurring in real

    time todayis to make it as easy or a working mother

    o our to review class lectures on her iPod touch as it is

    or a serviceman in Iraq to complete the coursework and

    tests needed or his degree, working rom his barracks

    in Baghdad. Coppins technology strategy has won itnumerous awards and plaudits rom U.S. News and World

    Report, Network World, EDUCAUSE and others.

    INTEGRATED, IMMERSIVE LEARNING ENVIRONMENTS

    Leading colleges and universities are doing everything

    possible today to oer sophisticated, technology-rich

    environments that allow students to immerse themselves

    anywhere, anytime in interactive learning experiences. For

    the University o Connecticut School o Business and Coppin

    State University in Baltimore, a Fit or Purpose Network rom

    Avaya is making it possible.

    At the University o Connecticut School o Business, it

    means an inrastructure that can support essentially

    every application its student will use in businessromaccounting programs, to all the major nancial packages

    to a mock trading room. The school now has the scalability

    it needs or seven years, and achieved immediate

    savings in hardware costs, energy, backend updating

    and maintenance. According to Michael Verteeuille, Chie

    2http://www.avaya.com/usa/topics/avaya-data-networking/3Tolly Test Report: Avaya Data Network Solutions, Perormance, Resiliency and TCO Comparison to Cisco/HP ProCurve Across Network Classes, March, 2010

  • 8/6/2019 Issue1 Final

    5/14

    5

    Consolidating services: Where it makessense, Avaya consolidates services. The

    Avaya Secure Router 4134 can integrate

    WAN routing, LAN switching, voice

    gateway and a ull-eatured Unied

    Communications applications suite in

    a single device. Delivering solutions

    that require ewer devices provides

    consistently higher levels o reliability,

    perormance and eciency, ultimately

    resulting in lower capital and operating

    costs. See TCO Comparison (Figure 1).

    Energy eciency: Avayas industry-leading energy eciency delivers

    immediate operational savings by

    reducing the amount o power that is used

    to directly run a given set o equipment,

    and also indirectly due to the savings

    realized or reduced cooling requirements.

    See Power Consumption (Figure 2).

    Ecient Management Platorm: Rather

    than relying on multiple management

    systems that create a dispersed

    approach, Avayas network management

    consolidates voice, data, wired andwireless management systems in a single

    platorm that allows network managers totroubleshoot up to our times aster.

    Scalability or Future-Readiness

    No individual product can ever be cost-

    eectively uture-prooed. Instead a Fit or

    Purpose network takes advantage o a

    variety o innovationsvirtualized, multi-

    device platorms, streamlined architectures,

    automatic QoS and role-based securityto

    more eectively scale as business demands

    evolve and grow, maximizing capacity

    utilization, easily accommodating incremental

    growth and avoiding unnecessary and costlyork-lit upgrades.

    Virtualized multi-device platorms:

    Reducing the reliance on separate,

    dedicated platorms connected together

    in avor o a smaller number o multi-

    purpose hosts provides higher levels

    o capacity and throughput. The cost o

    premium, mission-specic hardware is

    avoided. New implementations are now

    simply sotware upgrades.

    Pay-as-you-grow scalable capacity: TheAvaya approach overcomes the limitations

    and constraints acing other systems(such as Token Sharing/Passing systems

    or basic Cascading) and instead makes

    it possible to scale proportionally as new

    switches or ports are added and as the

    requirement or more bandwidth grows.

    Automatic QoS: Avaya Automatic

    Quality-o-Service (QoS) seamlessly

    calibrates perormance on end user

    devices, servers, applications and

    Ethernet Switches, delivering consistent

    and optimized QoS that enables the

    network to better understand andrespond to latency-sensitive, high-priority

    trac in times o network congestion.

    WLAN: Current WLAN architectures

    whether Distributed/Fat AP

    implementations or Centralized/Thin

    AP congurationsare not adequate

    to support the expanded role WLANs will

    play in the overall enterprise network.

    Avaya unies wireless and wired

    networks using a next-generation, split

    plane architecture that oers the best

    solution or supporting the new role o802.11n-based WLANs and can achieve

    update les. And there are numerous other benets or

    ans, Canadiens sta and other stadium personnel: soon

    the public will even be able to locate specic acilities

    like concessions, boutiques, and an zones with a

    positioning location agent on their smart phones, using

    the WiFi network. The bottom line: the team that haswon more Stanley Cups than any other ranchise now

    has a permanent power play when it comes to mobile,

    multimedia communications.

    A POWERPLAY IN MOBILE, MULTIMEDIA

    COMMUNICATIONS

    Some o the greatest players in the history o the game

    have cut the ice at the Bell Centre in Quebec, home to

    the Montreal Canadiens. So when the stadium decided

    it needed better mobile, multimedia communications, it

    wanted a solution that was in its league: an Avaya Fit or

    Purpose voice and data network emerged as the standoutchoice. Now, thanks to the power and scalability o the

    Avaya WLAN, greeters can use wireless handheld scanners

    to validate tickets; ood and beverage vendors can use

    WiFi point-o-sale devices; reporters and photographers

    can use the new wireless service to securely upload and

  • 8/6/2019 Issue1 Final

    6/14

    6

    up to a 55% reduction in hardware costs

    compared to the market share leader.

    Role-based Access Control: Many

    organizations have created a siloed

    security architecture with separate

    directory stores and policies or the wired

    network, WLAN, VPN, etc. Avaya simplies

    network identity management across

    the enterprise, providing a consistent,

    centralized, role-based access that can

    be implemented and managed on a ull

    network-wide basis.

    5. THE FUTURE OF FIT FOR PURPOSENETWORKS

    The era o real-time communications and

    collaboration is a game-changer in the

    world o enterprise network architecture. The

    lessons rom existing networkssuch as

    the Vancouver 2010 Olympic and Paralympic

    Winter Games, with its 20-old increase in

    bandwidth requirements (Avaya was the

    Ocial Converged Network Equipment

    Supplier or the Games)clearly show

    that the change is not simply incremental.

    The demands o resiliency, eciency

    and scalability require a paradigm shit

    in network design. The need to careully

    scrutinize data networking purchases

    to maximize ROI and limit TCO will only

    increase with the growing UC demand.

    Enterprises that are looking to cost-eectively leverage the latest innovations to

    drive ongoing productivity and competitive

    dierentiation need to take account o the

    new environment.

    In short, the time is now or networks that

    are truly Fit or Purpose.

    Learn More

    Avayas vision or the enterpriseplug

    & play communicationscalls or a new

    level o synergy between people; the

    collaborative, real-time applications that t

    use; and the underlying, enabling networ

    The goal is to deliver a consistent, enterpr

    class experience, one that supports conte

    aware collaboration that is always-on,

    simplied and virtualized. A key building

    block or this vision is the Fit or Purpose

    network. As real-time communications

    continue the evolution to IP, the data netwbecomes totally integrated into the deliver

    communications enabled business service

    in addition to transporting non-real-time,

    critical business applications. For more

    inormation on the intelligent, cost-saving

    capabilities o Avaya Fit or Purpose netwo

    solutions, visit www.avaya.com.

    Source: A

    Sources: Tolly Test Reports: Converged Data Network Solution, Evaluation o Energy Consumption and Projected Costs or a Converged LAN Campus, DataCenter and WAN, March 2010; Ino-Tech: Avaya vs. Cisco: A Comparison o LAN and WAN Inrastructure, April 2010

    FIGURE 1 FIGURE 2

  • 8/6/2019 Issue1 Final

    7/14

    7

    DELIVERING THE FIRST-EVERCONVERGED OLYMPIC ANDPARALYMPIC GAMES NETWORK

    Avaya teamed with the Vancouver

    Organizing Committee or the 2010 Olympic

    and Paralympic Winter Games (VANOC)

    and Bell Canada to help deliver a fawless

    Games experience or the XXI Olympic and

    Paralympic Winter Games.

    2010 Winter Games

    Avaya was chosen by VANOC as their

    Ocial Converged Network Equipment

    Supplier or the 2010 Olympic and

    Paralympic Winter Games. Avaya supplied

    network communications equipment both

    directly to VANOC and to Bell Canada, the

    exclusive Telecommunications Provider or

    the 2010 Winter Games, to build the rst all-

    IP converged network at an Olympic Games.

    The scenario

    On February 12, 2010, some 5,000 athletes

    and ocials rom more than 80 countries

    were on hand at BC Place Stadium in

    Vancouver, British Columbia, Canada or

    the opening ceremonies o the XXI Olympic

    Winter Games.

    Over the course o the next 17 days, roughly

    10,000 media representatives converged on

    the Olympic venues to cover seven sports

    and 15 disciplines. Over a million guests

    arrived at nine competition venues and three

    billion television viewers rom 160 countries

    were enthralled by the spectacle.

    A month later, ten days o Paralympic

    Games events were held.

    Provisioning a communications inrastructureor such an event is quite clearly an enormous

    challenge. The key to delivering seamless

    communications or this high-prole event?

    Convergence.

    Says Justin Webb, Bell Canadas VP or

    Olympic Services: When you look at the

    communications requirements o the Olympic

    and Paralympic Games, where you have a

    massive amount o activity in a very short

    period o time, the simplest way o doing

    this is over a converged IP inrastructure. So

    as much as possible, every service oeringthat Bell provides will be put over IP hence

    were calling them the All-IP Winter Games.

    Completing the jigsaw

    In terms o the scope and scale o this

    undertaking, VANOCs vice president o

    technical inrastructure, Andy Platten, helps

    put things into perspective:

    I used to work in IT or a very large bank.

    The inrastructure that were building or the

    Games is roughly the same complexity and

    size as what we had deployed there.

    But what were doing here or the Games is

    like opening all the branches on the same

    day, with all the systems working. And it has

    to run perectly. Customer transactions cant

    go wrong, and we have to balance every

    night.

    Its an interesting challenge.

    Finding a trusted vendor

    For both VANOC and Bell Canada, preparing

    to deliver this array o services has o courseentailed extremely meticulous preparation,

    beginning with the selection o their partners.

    We ran a highly competitive process, says

    Webb, but what we did that was dierent

    was that we went to the market or a partner.

    We needed a partner because we dont

    yet have nal requirements. We know the

    essence o what were going to deliver, but

    we dont know exactly, to the last detail,what the nal requirements will be.

    It was a very rigorous process. And quite

    rankly, Avaya came orward and was

    signicantly more committed to partnership,

    more committed to the spirit o the Games

    and more committed to being fexible in

    delivering the solution, even though we

    havent yet settled on the nal details.

    O VANOCs decision to partner with Avaya,

    Platten says, What drew us to Avaya was

    the level o commitment they were preparedto make in terms o how engaged they

    would be as an organization and the caliber

    o the people they were bringing in.

    In putting on the Games, says Platten,

    only ve percent o the sta comes rom

    VANOC 95 percent comes rom partners

    and volunteers who need to come together

    to orm one team. We needed a vendor we

    could put our aith into to be a true partner.

    We also liked the act that they gave us

    the opportunity to have one end-to-endnetwork. Bell Canada had already selected

    them as a partner or wide area networking

    and voice. And the jigsaw wasnt complete

    until we selected the supplier or the LAN

    equipment. Having Avaya be the partner

    in that piece o the jigsaw gave us our total

    solution. It helps us with interoperability and

    it gives us accountability: one team to go to

    resolve any issues.

    A big win

    At the heart o the 2010 Winter Games

    network was the Avaya Ethernet RoutingSwitch portolio or a resilient, high-

    bandwidth core network that connects all

    venues and supporting sites. The Avaya

    Ethernet Routing Switch portolio with Power

    over Ethernet converged wired and wireless

    communications across all locations.

    Says Justin Webb, We were very impressed

    with the carrier-grade attributes o Avayas

  • 8/6/2019 Issue1 Final

    8/14

    8

    Ethernet Routing Switches, and the capacities

    and densities they oer will save us money

    by allowing us to deploy ewer boxes.

    Mobility is another key aspect o providing

    customer service or the Games. Says Webb:

    How ast you can get access to inormation

    is important or the customers well be

    serving at the Games. When a journalist is on

    the bus traveling rom an event in Whistler toone in Vancouver, hes going to want network

    access to be able to upload his articles.

    To meet this requirement within each Games

    venue, a Avaya Wi-Fi network provided

    anytime, anywhere access to inormation.

    In addition, CDMA cellular acilities were

    expanded to deal with the surge in customer

    trac expected or the 2010 Games.

    Bell Canada also oered network-based

    VoIP within all Games sites with the Avaya

    Communication Server 2000 Compact and

    Avaya IP Phone 1100 Series sets. In act, the

    Vancouver 2010 Olympic and Paralympic

    Winter Games were the rst Games to

    exclusively use voice over IP or all event

    locations, providing increased customer

    service through the ability to provide a single

    phone number that can be used seamlesslyacross traditional handsets and computers/

    laptops.

    As Platten mentioned, securing the network

    end-to-end is another strong requirement

    or an event with such a high prole. Using

    the Avaya Secure Network Access product in

    conjunction with Avaya Switched Firewalls

    Intrusion Detection Systems and VPN

    concentrators ensures the network is not

    only protected rom intentional threats ro

    the outside but also unintentional threats

    rom inside. It ensures that devices that

    are being brought rom all over the world

    by athletes, ocials and the media are

    rigorously checked to ensure they are ree

    rom malware and comply with the rigidsecurity policies dened or the 2010 Winte

    Games.

    Perhaps the most signicant and high prole example o a

    Fit or Purpose network was the one Avaya delivered or the

    2010 Olympic and Paralympic Winter Games in Vancouver,

    Canadathe rst all IP network in Olympic history.

    Like the 5,000 Olympic athleteseach one a world

    competitor with years o training in his or her sport

    Avaya (as the Ocial Converged Network Equipment

    Supplier) brought its in-depth experience in real-time

    communications to the design and implementation o an

    end-to-end network or 15 Games venues and numerousnon-competition sites. These included two data centers,

    two media centers, two athlete villages, two ceremonial

    sites and the Games headquarters. The network had to

    meet Olympic requirements or 100% uptime in a converged

    voice/data, wired/WiFi solution that ultimately handled:

    4.9B bytes of data traversing the Internet a 20%

    increase over the 2008 Summer Games

    1.1 trillion web page downloads

    10Gb of bandwidth to every venuea 200% increase

    over the 2008 Summer Games

    1.1B data packets over Bells network

    30,000,000 MB: Total mobile data count

    65,000,000 mobile text messages sent and received

    750,000 calls made from 6,000 landline VoIP phones

    The scalability challenge: The Avaya network or the

    Vancouver 2010 Olympic and Paralympic Games

    demonstrated the unprecedented scalability challengesacing IT managers as more real-time communications and

    applications are brought on to the network:

    The Avaya network was designed for unanticipated

    surge capabilities equivalent to roughly three Super Bowl

    Sundays per day or 17 days.

    Overall, in the 1.5 years between the 2008 Summer

    Olympic Games and the 2010 Winter Games, the

    bandwidth requirements grew twenty old.

    A FIT FOR PURPOSE NETWORK FOR THE WHOLE WORLD TO SEE

  • 8/6/2019 Issue1 Final

    9/14

    9

    We were looking or a partner who will

    be here and who will help us plan, says

    Platten in summation. The act that Avaya

    will bring their expertise to the table and their

    commitment to deliver, on time, whatever we

    design and do whatever it needs to make

    sure it works seamlessly is a big relie to

    us.

    Learn More

    For more inormation on the intelligent, cost-

    saving capabilities o Avaya Fit or Purpose

    network solutions, visit www.avaya.com.

    Source: Avaya

    I used to work in IT or

    a very large bank. The

    inrastructure that were

    building or the Games

    is roughly the same

    complexity and size as

    what we had deployed

    there. But what were

    doing here or the Gamesis like opening all the

    branches on the same

    day, with all the systems

    working. And it has to

    run perectly. Customer

    transactions cant go

    wrong, and we have to

    balance every night. Andy Platten, VP o technicalinrastructure, VANOC

    2010 WINTER GAMES A TRULY HYPERCONNECTED EVENT

    The inrastructure supporting the 2010 Winter Games will not only have to

    support over a million individuals it will also have to accommodate a

    myriad o dierent devices, which is why delivering a network designed

    or capacity, fexibility, reliability and simplicity is o the utmost importance.

    15,000 VoIP phone and fax lines

    7,000 mobile phones and 2,000 push-to-talk

    5,000 radios

    4,000 TV drops

    500 Wireless Access Points

    BEHIND THE BELL ALL IP GAMES

    The Vancouver 2010 Olympic and Paralympic Winter Games were the

    rst Games run on an all-IP converged network. Previous Games have

    eatured at least our purpose- built networks or voice, data, video and

    broadcast services.

    WHY IP?

    Increased fexibility: Throughout such a dynamic event, the network

    needs to be able to quickly adapt to changes.

    Increased customer satisaction: Customers will receive a single

    phone number that will work seamlessly across traditional handsets

    and their computers/laptops.

    Increased cost-eectiveness: Delivering a fawless Games

    experience is the baseline. Doing it more cost eectively? Even better.

    Avaya came orward

    and was signiicantly

    more committed to

    partnership, more

    committed to the spirit

    o the Games and morecommitted to being

    lexible in delivering the

    solution...

    Justin Webb, VP o Olympic Services, BellCanada

  • 8/6/2019 Issue1 Final

    10/14

    10

    From the Gartner Files:

    Top 10 Tips to Prepare Your Networkfor Unified Communications andCollaboration

    This documents details the steps required to ensure that appropriate

    inrastructure is in place to support unied communications andcollaboration (UCC) rollouts. Lack o appropriate inrastructure will

    result in user dissatisaction and, potentially, ailure o the overall UCC

    initiative.

    Key Findings

    Most existing networks are ill-suited to support UCC, particularly

    ad hoc multipoint videoconerencing. UCC applications introduce

    new protocols and network trac patterns with increased

    dependence on quality o service.

    Required upgrades involve a broad range of systems, from

    Multiprotocol Label Switching (MPLS) backbones to accessrouters, as well as the introduction o new components such as

    WAN optimization controllers and Session Initiation Protocol (SIP

    rewalls. For more advanced security requirements, consider

    session border controllers.

    As acceptance of UCC grows, dependence on the underlying

    inrastructure will increase. Ongoing evaluation o the business

    impact o inrastructure ailure must be done to ensure that an

    appropriate level o investment is maintained to support these

    systems.

    Joe Skorupa:

    Joe Skorupa is a research vice president in Gartner Research, where he is part o the Data Center Transormation

    and Security Research unit. His research ocuses on maximizing application perormance via advanced networking

    technologies, including Layer 4-7 switching, and WAN optimization. He also covers data center network convergenc

    abric-based inrastructure and delivery o rich media, including video within the enterprise. In addition, Mr. Skorupa

    advises vendors on product, market and partnership strategies. Through extensive collaboration, Mr. Skorupas

    research extends to intersections between networking and adjacent research areas, including storage, servers and

    data center optimization. The research covers topics including Fibre Channel over Ethernet (FCoE), I/O virtualization,

    and emerging data center architectures, such as cloud computing and abric-based computers.

    Mr. Skorupa has more than 25 years o experience in enterprise, service provider and storage networking. He has hel

    a variety o engineering, marketing and analyst roles at companies such as FORE Systems, Bytex, Motorola and RHK.

    Debra Curtis:

    Debra Curtis is a research vice president in the Gartner IT Operations Management group. She is responsible or the

    research core topics o network management, event management, business service management and IT service

    portolio management. She ocuses on best practices or IT operations management, improving IT management

    process maturity and end-to-end IT service management.

    Prior to joining Gartner, Ms. Curtis held product marketing and executive management positions at DeskTalk

    Systems and AXON Networks, both vendors specializing in network perormance management. Ms. Curtis also

    spent 15 years at Digital Equipment Corporation, holding positions in product marketing, marketing planning,

    international marketing, sotware consulting and technical support.

    Lawrence Orans:

    Lawrence Orans is a research director in Gartners Research organization. His research ocuses on the integration o

    security within internal networks, with a particular emphasis on network access control, VoIP and content ltering.

    Prior to joining Gartner in 1997, Mr. Orans held systems engineering positions at IBM, 3Com and Bay Networks (via

    Wellfeet Communications), where he worked with enterprises on designing and implementing secure network

    inrastructures. In total, he has more than 20 years o experience in the industry.

  • 8/6/2019 Issue1 Final

    11/14

    11

    Investment in staff training and in new design and troubleshootingtools is required to ensure a successul design, deployment and

    operation o UCC systems.

    Recommendations

    Involve the network and security teams early in UCC planning,

    pilots and deployments.

    Set proper expectations that during periods of peak demand some

    services, particularly video-based services, may have to be limited

    due to network bandwidth constraints.

    UCC deployment can signicantly increase capital and operational

    costs or networking and security, so include these additional costsin the overall return on investment (ROI) calculation or UCC.

    ANALYSIS

    Many organizations will begin initial deployments o enterprisewide

    UCC capabilities in 2010. While much attention will be paid to Internet

    Protocol (IP) telephony call managers, video, shared work spaces and

    presence, most organizations will not devote adequate planning and

    resources to the required network inrastructure upgrades. However,

    without these upgrades end-user productivity and satisaction

    regarding the UCC tools will not meet expectations.

    Gartner believes that UCC deployment must be addressed in the

    broader context o overall network re-engineering. UCC applicationschange network trac patterns rom hub and spoke to mesh as

    branches communicate directly with one another. Trac loads

    change rom non-real-time to latency and bandwidth-sensitive real-

    time communications, including ad hoc multipoint videoconerencing.

    Additionally, new protocols and servers are introduced into the

    environment, complicating design and operations.

    The impact o ad hoc high bandwidth services (such as point

    to point and multipoint video) can negatively aect overall UCC

    services. Advanced scheduling combined with call admission control

    (denying new sessions when at capacity) can help. But it is critical

    to set expectations that not all services will be available under any

    conditions and then to train the users on how to all back to voice-onlycalls.

    Over time, demand or rich collaboration with employees that are o

    the corporate network or with business partners will develop. This

    will drive an increase in Internet access bandwidth and will introduce

    an additional set o interoperability and security concerns that are

    beyond the scope o this note.

    As with any IT investment, the level o investment must match the

    impact o the system on the business. As UCC systems are an

    emerging technology, the impact on the business may changeas end-user acceptance and dependency evolves. Ongoing risk

    assessment and involvement o business leaders is critical to ensure

    that the appropriate investments in inrastructure and stang are

    maintained.

    Assure UCC Perormance

    UCC introduces new trac patterns and protocols to the network.

    While the signaling trac tends to be many to one (endpoints to

    limited number o SIP servers), the session trac will oten travel

    directly between endpoints. As a result, a ull mesh quality o service

    (QoS) implementation is required. Because many applications use

    HTTP or HTTPS as their transport, the QoS mechanisms must include

    deep packet examination to ensure that critical trac is protected.

    Keep in mind that video consumes approximately 10 times the

    bandwidth o a voice call. As UCC gains acceptance, its use will

    increase and use patterns will emerge. To avoid an ever escalating

    demand or more bandwidth, it will be necessary to monitor ongoing

    use, working with teams that have regularly scheduled meetings

    to alter the meeting times, or adjusting the available tools during

    especially busy times.

    1. I you havent already done so, install redundant WAN links

    between critical sites to ensure a network outage does not disrupt

    communications between these sites. Keep in mind that relatively

    small sites may be critical to important projects. For these sitesthe redundant links may be provided by Internet virtual private

    networks (VPNs), although the perormance o these links may be

    less predictable.

    2. Implement QoS enorcement in the edge routers to protect

    critical SIP, voice over IP (VoIP) and video trac, as well as other

    mission-critical trac. Many companies connect to small branch

    oces via Internet VPNs, which provide best-eort transport

    with unpredictable perormance. This may be inadequate or

    UCC trac. I so, these links will have to be upgraded to MPLS

    connections, which will urther increase the operational expense

    or UCC.

    3. Consider WAN optimization controllers (WOCs) that provide deep

    packet examination to inspect and mark the trac. In some cases,

    asymmetrical examination and enorcement may be adequate. In

    other cases, symmetric implementations may be required. Keep

    in mind that i any o the trac is encrypted by IPsec, the overall

    eectiveness o the network may not be able to be guaranteed.

    4. I bandwidth is scarce and expensive consider a combination o

    the ollowing:

  • 8/6/2019 Issue1 Final

    12/14

    12

    1. Ofoad less critical trac (Web surng, le transers, e-mail)to an Internet VPN to preserve bandwidth on the QoS-enabled

    network or perormance-critical UCC trac. This approach can

    also provide some degree o resiliency should the primary link

    ail.

    2. Use advanced platorm application delivery controllers (AP

    ADCs) and WOCs to ofoad busy links where more bandwidth

    is not practical. AP ADCs can improve the perormance o

    interactive applications by 50% and can improve the response

    time or Microsot SharePoint by as much as 50% without

    requiring an appliance in the remote oce. WOCs can reduce

    response times by 50% to 90% and can reduce peak bandwidth

    requirements or non-video/voice applications by as much as90%, although average improvements are typically closer to

    50%.

    In many cases, a hybrid approach will be the most cost-eective

    approach with some sites served by AP ADCs while others require

    a combination o AP ADCs and WOCs. I multipoint voice and video

    calls are expected to be common, bandwidth upgrades will probably

    be required. I edge QoS and ofoad prove inadequate you may

    have to upgrade your MPLS services to include QoS support, which

    can be expensive. This will also require additional trac engineering

    work across the backbone. Be sure that the cost o this upgrade is

    considered, i not explicitly included, in the overall budget or the UCC

    implementation.

    Adequately Size and Secure Your Inrastructure

    UCC architectures are dierent rom traditional productivity

    applications. Most activities depend on one set o servers to set up

    communications and others to provide shared tools, and both sets

    o devices represent potential perormance bottlenecks and single

    points o ailure. End-user satisaction with UCC implementations is

    directly linked to the availability and perormance o those systems.

    5. Ensure that multiparty services such as shared whiteboard or

    multipoint voice and video bridging have adequate capacity or

    anticipated use. While raw server perormance is a component

    o this requirement, network perormance can also aect overall

    system perormance. Take care to ensure that servers have

    adequate network bandwidth and that servers are located to

    minimize latency between end users. This may require regional

    servers. In cases where collaborators are distributed globally,

    latency may be an ongoing issue. In these cases it is important to

    set proper expectations during initial system rollout.

    I at all possible, select UCC products that support the call admissicontrol concept, so that, or example, i a site can only support 20

    video calls then the UCC system will block the 21st. It will then be

    important to coordinate capacity management between the netwo

    management systems and the UCC platorm.

    6. Implement redundant servers (o ten as distributed redundant

    pairs) with load balancing or all critical unctions. As users

    come to depend upon UCC systems to perorm their everyday

    activities, unplanned (and even planned) outages can result in

    signicant disruptions to the business, loss o productivity and u

    dissatisaction. O course, the investment in redundant systems

    must match the economic impact o the outage. Running UCC

    services as virtual machines can reduce the cost o redundancysharing replacement pools with other virtualized applications.

    7. While attacks against unied communications systems have

    been ew and ar between, UCC systems can be easily attacked

    and rendered inoperable, particularly by insiders. The level o

    protection or a UCC system should match its level o criticality. T

    main risk to a UCC system is that it could be disabled by an attac

    (or example, denial o service) on one o the UCC servers.

    Most UCC systems are based on the SIP signaling protocol, and a

    number o ree tools are available on the Internet to launch SIP-bas

    attacks. Also, SIP vulnerabilities are being exposed at a aster rate

    than with other signaling protocols, due in large part to the act thaopen source SIP implementations are widely available to researche

    and attackers.

    To mitigate these risks, protect the UCC servers with internal rewa

    that provide stateul inspection o SIP (or whichever signaling

    protocol is being used). Organizations that are more security

    conscious may choose to complement the rewalls with an intrusio

    prevention system that is capable o preventing signaling attacks o

    implementing a session border controller (SBC).

    Alternatively, SBCs can be used to reduce SIP-based denial o servi

    (DoS) threats that originate rom within and outside an organization

    and provide interoperability with various versions o SIP being used

    by service providers and enterprises. The complementary unctions

    session management and session border control improve enterpri

    communications security, UCC application deployment, operationa

    eciency and reliability.

  • 8/6/2019 Issue1 Final

    13/14

    13

    Match Your Tools and Training to the New Environment

    UCC trac will share the network with existing trac, making

    monitoring and troubleshooting more dicult and rapid problem

    resolution more important.

    8. Consider general purpose network perormance monitoring

    tools, rather than point products that can only monitor UCC trac.

    General purpose network monitoring tools are able to monitor

    UCC trac, but will evaluate it relative to the rest o the business

    applications that are also using the network bandwidth. UCC

    should be managed in the context o the broader enterprise

    network, not treated as a separate stovepipe. Point products only

    look at UCC perormance, whereas general purpose tools monitorthe quality o service or all network trac, not just UCC. Beore

    adding UCC trac to the data network, its vital to get a baseline o

    current network service levels and business application transaction

    perormance, in order to be able to report any change in service

    quality ater the UCC deployment.

    9. Employ network emulation, simulation and modeling tools to

    determine i network latency and bandwidth limitations will

    negatively impact the end user perception o UCC applications. A

    network assessment should be perormed not only to determine

    whether UCC trac can be carried with adequate service

    quality, but also to predict the impact o UCC trac on business

    applications that already run on the network.

    As part o the capacity planning process, these tools can beused to help ensure optimal network design to support UCC and

    current business applications, as well as to reliably predict uture

    perormance based on an understanding o business demands,

    including additional users and new applications. Follow the planning

    process with a limited scale pilot to ensure assumptions are valid and

    to amiliarize the team with the new products and procedures.

    10. Upgrade network engineering and operations sta tools and

    training. UCC is an implementation o multiple products rom

    multiple vendors. I the engineering and operations sta is

    unable to adequately design, deploy, operate and support the

    UCC environment, the project will quickly be seen as a ailure

    by end users and top management. The introduction o SIP andother new protocols and technologies into the network, as well

    as the introduction o new trac patterns and loads, may require

    additional sta training. New design and troubleshooting tools

    may also be required, each with their own training requirements.

    Gartner RAS Core Research Note G00201607, Joe Skorupa, Debra Curtis,

    Lawrence Orans, 6 July 2010

  • 8/6/2019 Issue1 Final

    14/14

    14

    Avaya is a global leader in enterprise communications systems. The company provides unied communications,contact centers, and related services directly and through its channel partners to leading businesses and

    organizations around the world. Enterprises o all sizes depend on Avaya or state-o-the-art communications

    that improve eciency, collaboration, customer service and competitiveness.

    For more inormation please visit www.avaya.com.

    About Avaya