building high leverage networks

Building HigH leverage networksDelivering ApplicAtion Fluency For your network Service oFFeringS white pAper

Contents

the growing network infrastructure crisis / 1

real-time Applications push legacy networks to their limit / 1

Smart Devices increase the pressure / 1

virtualization Squeezes the network from the inside / 2

Application Fluency: A new Approach for network Designs / 2

high-Quality user experience / 3

lower network Administration costs / 3

Better return on investment (roi) / 4

Building the high leverage network using Application Fluency / 4

resilient Architecture / 4

Simplified network / 4

virtualization / 5

transparent resiliency / 6

embedded Security / 6

Automatic control: the intelligent network / 6

profile Directed / 6

Autosensing / 6

policy Driven / 7

end-to-end control / 7

Streamlined operations / 7

Automatic provisioning / 7

unified Management / 7

open Standards / 7

energy efficient / 8

Solutions enabled by the Application Fluent network / 8

Managed Access Services / 8

Service Aware networks / 9

the Data center / 10

Managed communication Service networks / 11

Application Fluency: th e next generation of high leverage network infrastructure / 11

why Alcatel-lucent omniSwitch portfolio? / 12

Building high leverage networks: Delivering Application Fluency for your network Service offeringsalCatel-luCent wHite PaPer

1

tHe growing network infrastruCture Crisis Today’s network infrastructure is facing an unprecedented set of challenges — challenges that threaten its ability to meet rising user expectations. At the same time, these challenges create opportunities for operators that choose to meet them head-on. This paper will show how your network infrastructure can respond effectively to a shifting climate to create competitive advantage, drive customer loyalty and minimize cost.

But first, let’s take a look at some trends that are putting pressure on the network: mission-critical and real-time applications, smart devices and virtualization.

real-time applications Push legacy networks to their limit Delivering real-time applications is a top-of-mind issue for business managers and operators alike. Voice over IP (VoIP), video, social media, customer service, and collaboration suites are fast becoming essential tools for users to interact effectively and stay engaged in this digital age. These next-generation applications require levels of performance from the network infrastructure far beyond the needs of yesterday’s networks in terms of throughput, latency and application awareness.

In response, operators must build a network that has the right mix of resources to provide a satisfac-tory experience to the end users. Traditionally, operators have played a limited role in optimizing application delivery to the end users: they simply provided the internet connectivity. Applications and consumers of those applications are either co-located or connected over best-effort networks provided by the operators. In this environment, architects designed the network with enough raw bandwidth to meet peak application delivery requirements within a location and enable limited use over the best-effort network. This approach — known as over-provisioning and restricted access —provided adequate performance, but clearly in limited scenarios and with limited flexibility.

To complicate matters, many networks were already running close to their practical limits in terms of complexity, performance and reliability. The reasons are varied; in some cases, networks grew organically over years or even decades with little standardization or central control, resulting in incompatible components, protocols and management tools. Portions of the network frequently were modified to meet the needs of specific new service offerings and applications, adding complex-ity and driving up operating costs. Mergers and acquisitions created silos of incompatible technology coexisting within a single network. The business impact of these trends is significant: a rigid infrastructure unable to deliver new service offerings and provide a rich user experience for the new wave of mission-critical, real-time services and applications.

smart devices increase the PressureThe stress on the network is only going to increase, thanks to the rapid adoption of always-connected devices and bring your own device (BYOD) culture. Network operators are struggling to keep up with bandwidth demand and provisioning requirements from these always-on varied smart devices. BYOD culture is also adding tremendous challenges to keep the sanctity of the network while providing the rich and versatile user experience.

These new rich-media devices do more than eat up bandwidth: they make it increasingly difficult for network managers to predict bandwidth consumption. In the absence of a clear picture of network traffic patterns, the traditional practice of setting static priority levels for specific applica-tions no longer works. Network planners need a fresh approach to support the new generation of connected devices.


2

virtualization squeezes the network from the insideUnlike the previous two trends — real-time applications and smart devices — virtualization is an internally-generated pressure. However, its impact on the overall network design can be just as dramatic. To increase agility, enable new service offerings, optimize resources and reduce costs, networks are designed with virtualizing servers, storage and networks in ever-larger numbers. The relentless push to virtualize the data center to support the growing use of cloud-based service offerings increases baseline network traffic dramatically — and rewrites the rules of network design and operation and is particularly challenging when delivered as a cloud offering.

Virtualization is also moving beyond the data center. Network Infrastructure and IT groups are deploying network appliances to deliver specialized services such as intrusion protection, traffic management and faster service enablement. Virtualized desktops are gaining in popularity due to their reduced support costs and better security. These and other applications of virtualization will create additional demand for raw bandwidth and require the network to automatically react to the needs of virtualized systems — something legacy networks were never designed to accommodate.

For the first time, we are seeing the true marriage of WAN infrastructure and the enterprise network due to the trends we discussed above. Network infrastructure is becoming the melting pot for all the technology trends like mobility, social, virtualization and globalization, where physical boundaries are no longer relevant to the overall application performances and end-to-end user experiences.

The time-tested solution of adding more and more raw bandwidth will not work. What’s needed is a new resilient network architecture which is designed to deliver a consistent user experience irrespec-tive of location, access media, device and application. A network architecture which is able to adapt and respond to the varied application needs in real-time to ensure user quality experience.

aPPliCation fluenCy: a new aPProaCH for network designs Application fluency1 represents a unique approach to networking design. In Alcatel-Lucent’s view, the application fluency based network (also called Application Fluent Network) possesses broad knowledge of network devices, end terminals and the applications to which they are connecting. Most importantly, the application fluent network understands the context of the conversation between device and application — and makes decisions based on that understanding.

Alcatel-Lucent’s application fluent network is based upon a resilient architecture with streamlined operations and automatic control as shown in Figure 1.

Resilient Architecture: A simplified, single IP network with built-in security.

Streamlined Operations: Reduced complexity through end-to-end network visibility, automated provisioning and integrated troubleshooting tools.

Automatic Control: High-quality mission-critical and real-time application delivery with unique dynamic tuning of network performance.

1 term coined by gartner. .


3

Figure 1: Application Fluent Network

The application fluent network brings significant benefits to the operators including a high-quality user experience, lower network administration costs and a better return on investment (ROI).

High-Quality user experienceThe application fluent network allows network managers to offer a high-quality user experience with controls that automatically ensure service levels, avoid packet loss and provide security such as:

• Dynamic tuning of network performance. The application fluent network automatically adapts to maintain service levels for individual conversations based upon context and business priorities.

• Automaticrecoveryfromnetworkfaults.The application fluent network recovers rapidly from network element and link failures, avoiding any perceptible service interruptions.

• Embeddedsecurity.Security features in the application fluent network operate transparently to protect users and their conversations from threats.

lower network administration CostsThe application fluent network eases the burden on network operators with built-in features for automated provisioning, integrated management and end-to-end service level agreement (SLA) of the network including:

• Low-touch operations. The application fluent network manages provisioning of network elements with minimal skilled human intervention. The use of network and access policies ensures consistent configurations across the entire network, reducing time spent tweaking the configuration of individual network elements.

• Convergedmanagement.In the application fluent network, a consolidated management tool provides a single pane of glass for monitoring and managing the network. It also allows administrators to manage the network infrastructure as a single fabric rather than individual components. Operators spend less time switching between programs and screens, and more time on productive work.

Architecture

OperationsControl


4

Better return on investment (roi)The application fluent network both minimizes the initial capital investment and lowers operational costs, which allows the network operator to recoup its investment faster through:

• Flatter network architecture. By eliminating an entire network layer, the application fluent network reduces the number of network elements and hierarchy.

• Networkvirtualization.The application fluent network virtualizes network resources, resulting in optimized utilization, lower capital requirements and higher ROI.

• Greendesign.The new generation of energy-efficient high density network elements used in the application fluent network reduces expenses for space, power and cooling.

Building tHe HigH leverage network using aPPliCation fluenCyThe High Leverage Network (HLN) is an all-IP infrastructure that delivers very high bandwidth at lowest cost per bit, along with a richer experience for residential, business and mobile customers, helping operators to monetize their networks more effectively. Application fluency can be identified by a number of key properties that are required to deliver the desired user experience: architecture, adaptability and management; this is shown in Figure 2 and described in more detail below.

Figure 2: Key properties of the application fluent network

resilient arCHiteCture The architecture of an application fluent network can reduce complexity, cut costs and provide operators more flexibility. These are enabled through:

• Alow-latencynetworkdeliveredviaasimplifiedarchitecture

• Anoptimizedutilizationofnetworkresourcesachievedbyvirtualization

• Aresilientnetworkenabledbyminimizedre-convergencetimes

• Theremovalofsecuritycomplexitywithusercentricpolicies

In the remainder of this section we will explore each of these aspects of an application fluent network in more detail.

simplified network The application fluent network eliminates the explicit multilayer hierarchy of the network and builds a simple, three-tier network that can support residential, business and backhaul subscribers (see Figure 3). The network elements used in the application fluent network lessen complexity in

ResilientArchitecture

• Simplified Network• Virtualized• Resilient• Embedded Security

Automatic Control

StreamlinedOperations

• Profile Directed• Autosensing• Policy Driven• End-to-end control

• Automated Provisioning• Unified Management• Open standards

• Low Power


5

the aggregation layer because they feature wire-rate and non-blocking capacity as well as a high-port density, along with integrated management. The removal of the complexity inside the aggregation layer is also critical for achieving a lower latency network.

Figure 3: Simplifying the network topology

virtualizationBy virtualizing network elements and link management, the application fluent network optimizes resource utilization, reduces operational complexity and increases scalability. The virtualization of both network elements and link management, referred to as Multi-chassis link aggregation (MC-LAG), is central to designing a simplified network architecture without requiring the complexity and inefficiency of spanning tree protocol (see Figure 4).

Figure 4: MC-LAG versus STP

One converged, optimized, scalable, multi-access network allowing dynamic service creation and delivery

Converged RAN

Converged Service Control

Converged wireline access

Converged metro and aggregation

Convergedbackbone

Converged edge

Converged service-aware network management

Radio access

Residential/enterprise gateways

VDSL/GPON/Ethernet

Ethernet

Mobilitymanagement

Sessioncontrol

Policymanagement

MPLS/IP/Optical

PhotonicSwitch

Servicerouters

Multi-screen servicesWeb and Enterprise 2.0

MC-LAG LAG

MC-LAG

LAG

MC-LAG

LAG

LAGSTPBlocked

MC-LAG and LAG Spanning Tree Protocol and LAG

MC-LAG increases switching bandwidthMC-LAG allows dual-homingMC-LAG change only impacts link

STP blocks linksSTP does not allow dual-homingSTP change impacts whole network


6

transparent resiliencyMission-critical and real-time applications require low latency and high availability, which can be compromised by hardware problems. The application fluent network recovers rapidly from network element and link failures — without impacting application traffic. Of course, all network switches have redundant power supplies, fans, management and fabric. Ethernet Ring Protection (ERP) also ensures quick re-convergence time in case of switch or link failure (see Figure 5) for geographically distributed networks.

Figure 5: Ethernet Ring Protection (ERP)

embedded securityBecause every network design has varied security requirements, the application fluent network includes a robust set of embedded security features. Built-in access policy, intrusion detection and prevention (IDP) offer additional layers of protection that supplement existing end point and server security solutions. Network security is based on subscriber profiles and policies. Network elements feature overload protection to prevent failure due to Denial of Service (DOS) attacks. Integrity checks can also be enabled to ensure only the right network elements are allowed to join the network.

automatiC Control: tHe intelligent networkEnsuring a high-quality user experience requires the ability to adjust key network parameters on-demand and in real time. This ability to make dynamic adjustments requires:

• Knowledgeaboutsubscribersandendpointsdeclaredinprofiles

• Theabilitytomonitoractualservicelevels

• Policiestodeterminewhenandhowtoadjustnetworkparameters

The application fluent network recognizes subscribers and end points automatically and assigns profiles that regulate access and bandwidth usage. The network is then continuously monitored, with policies informing the update of key network parameters that affect the end-user experience when needed. Each requirement is explored in more detail below.

Profile directed Profiles embedded within the network contain information about the requirements for SLA and security for conversations that will be carried by the network. In the application fluent network, profiles for subscribers, end points and applications are globally defined and embedded in the network elements. These profiles also can move with context of subscriber and end point.

autosensing Autosensing allows the application fluent network to treat individual flow and conversation within the network as follows:

MC-LAG

Aggregation(Ring or Mesh)

Core(IP or MPLS)


7

• Assigninitialallocationofnetworkresourcesbaseduponpriority.Aseachsubscriberandendpoint accesses the application fluent network, it is automatically recognized — regardless of location — and assigned the right profile.

• Dynamicallyadjusttreatmentoftrafficforaparticularfloworconversationwhentherequiredservice level is not being met. By monitoring network status events the application fluent network can determine the service level and take corrective action to meet specific SLAs.

Policy drivenA policy engine and rule set embedded in the application fluent network determines which actions to take for network events such as:

• Anendpoint,subscriberorapplicationrequestingaccesstothenetwork

• Afloworconversationthatisnotexperiencinganappropriatelevelofservice

• Amoverequestforasubscriberorvirtualmachine(inthedatacenter)

end-to-end Control In an ideal world, an operator would like to have end-to-end control of a flow or conversation for each subscriber or end-point. An application fluent network with dynamic profile and policy along with end-to-end OAM provides this capability to ensure SLAs to their subscribers.

streamlined oPerationsTo relieve the burden on operators and reduce operational costs, the application fluent network focuses on several essential needs:

• Providingautomationwhereverpossible,suchaswithprovisioning

• Consolidatingmanagementtools

• Ensuringoperationalconsistencybetweenswitchfeatures

• Designingequipmenttominimizepowerconsumption

In the remainder of this section we explore each of these needs in more detail.

automatic Provisioning The application fluent network includes automatic setup capabilities for switches and routers. When a new network switch is added to the application fluent network, it downloads the right configura-tion through the network. The switch becomes a functioning member of the network with minimal human intervention.

unified managementA common complaint by operators is the proliferation of tools for monitoring and managing the network infrastructure. The application fluent network helps relieve this problem by:

• Consolidatingmultiplenetworkmanagementfunctionsintoasinglemanagementtool.Operatorscan access this system remotely allowing them to monitor and control the network from any location, at any time.

• Enablingmanagementofthenetworkinfrastructureasvirtualizedsegments,(forexample,allswitches on one geographical location would be treated as one virtualized complex), rather than as individual network elements.

• Includingapplicationtrafficmonitoringtoolsthatallownetworkoperatorstocollectstatisticssuch as loss, jitter, latency, response time, and packet loss.

• Offeringthecapabilityofintegratingthird-partytoolsusingwell-documentedSNMPandCLIimplementation.


8

open standards Alcatel-Lucent is a big proponent of open standard-based implementation which provides flexibility to operators to run multivendor environments without any unwanted vendor lock-in. It also ensures higher interoperability levels, fewer lab trials and faster time to market.

energy efficientMore and more network operators are moving to green networks as a cost-saving measure as well as a good citizen initiative. The application fluent network minimizes power consumption through the selection of network components designed for energy efficiency. As a major contributor to network power consumption, all switches in the application fluent network have market-leading low power consumption characteristics.

solutions enaBled By tHe aPPliCation fluent network The application fluent network is suitable for the most demanding deployments inside service provider networks. Typical target deployments for the application fluent network include:

• Managedaccessservices

• Service-awarenetworks

• Thedatacenter

• Managedcommunicationservices

Each of these examples is described in more detail below.

managed access servicesEnterprises that rely on a service-provider WAN to deliver high-quality applications are no longer satisfied with simple packet delivery: they are demanding that providers commit to a level of service for these critical applications. To meet this need, service providers need to extend the SLA control from their core network all the way to the enterprise edge. They must also manage SLA & Quality of Experience (QoE) to the demarcation point at the customer premises.

Figure 6: Managed Access Service Networks

Service Edge Service

Aggregation Service Access CPE, CLE & Cellsite

DSL, GPON

Home, GW

EthernetVPN

IP VPN

MPLS

Access Node

BUSINESS

RESIDENTIAL

MOBILE

Medium andLarge Enterprise

SmallEnterprise

7210 SAS-M/X

OS6850/6855

OS6850/6855OS6850E

7210 SAS-M/X

7705 SAR-8

OS10K

OS9000E

7450 ESS

7750 SR

7750 SRc12

OA5510

7210 SAS-E/M

7210 SAS-E/M

7705 SAR-F

OS6250/6400

OA5510

InBuilding 6250/6400

Active Ethernet


9

Figure 6 shows a typical managed access service network based on the application fluent network. Alcatel-Lucent provides a comprehensive portfolio to address different segment of the network:

• Variedlastmileoptions

• Comprehensiveserviceaccesstoaddressmultipledeploymentoptions

• Cost-effectiveserviceaggregationbasedonEthernet

• Market-leadingIP-MPLScore

The key architectural features include:

• Singleinfrastructuretosupportbusiness,residentialandmobilesubscribers

• Seamlessnetworkintegrationtoprovideend-to-endvisibilityofapplicationdelivery

• Visibility,monitoringandtroubleshootingcapabilitieswithonemanagementplatform

• Abilitytotroubleshootavailablecircuitswithcarrier-basedtoolsthatmonitortheavailability,reliability and end-user experience

• RichQoEwithsimplifiednetworktopology

• Plug-and-playcapabilitiestoreducecomplexityanddeploymenttime

service aware networksService aware networks extend the geographical reach in public and/or private WAN infrastructures and are typically deployed by organizations or operators that have specific requirements for reliable, high-quality service delivery in rugged environments where security cannot be compromised — for example, in utilities, transportation and public safety organizations. These operators are under pressure to deliver additional services while reducing costs, which often requires significant modernization of the existing infrastructure. Important requirements for these operators include:

• Ahighlyscalableandreliablenetwork— including the ability to operate in harsh environments

• Integratedsecuritythatprotectsthenetworkfromattackandpreventsunauthorizedaccess

• Guaranteedservicelevelsforspecifictrafficbaseduponbusinesspriorities

• End-toendmanagementandvisibilityontheservicelevelsactuallydelivered

Figure 7: Service Aware Network

Node RedundancyIP/MPISBackbone

HardenedDevices

MC-LAG MPLS/VPLS DualHomed

ERP

5620 SAM for End-to-End Management

OmniSwitch6850

OmniSwitch900xE

OmniSwitch 10K

7750 SR 7750 SR

OmniSwitch6855

OmniSwitch6855

OmniSwitch 10K

ISSU

ISSU

ISSU

ISSU

ISSU

ISSU

ISSU

ISSU

OmniSwitch6850


10

The end-to-end application delivery requirements of service aware networks demand an MPLS back-bone as shown in Figure 7. Application fluency is achieved in this network with the implementation of Virtual Output Queuing (VOQ) to ensure high-quality application delivery by adding carrier-class QoS to the MPLS backbone. Additional important elements include carrier-class reliability features such as multi-chassis link aggregation, Ethernet ring protection and dual-homed link aggregation. For harsh environments, devices that can operate at high temperatures without requiring a fan for cooling are a must. Management is provided end-to-end with a common management platform for visibility and troubleshooting.

the data Center The data center is undergoing a transformation driven by the need to reduce costs and improve agility. At the core of this transformation is the consolidation of servers in the data center to reduce equipment, space, cooling and energy costs — all made possible by server virtualization technology. Continued transformation in the data center is expected with the eventual conver-gence of data and storage networks. This trend is accelerating due to the “consumerization of IT” with cloud-based services. Data center consolidation is driving a refresh of the switching fabric with the following requirements

• Simplifiednetworkdeliveringresilient,lowlatencyandlowpowerswitching

• Efficientdeliveryofinter-servertrafficaswellasclient-servertraffic

• Networksupportforautomaticmovementofvirtualapplicationsbetweenservers

• TheabilitytocarrybothdataandstorageinformationoverEthernetnetworks

• Disasterrecoverybetweendatacentersites

Figure 8: The data center

Alcatel-Lucent offers an award-winning and comprehensive data center inter-connect and data cen-ter switching solution. Within the data center, as shown in Figure 8, an application fluent network provides a two-tier resilient and low latency network architecture that supports server virtualization with automated virtual machine migration. It also provides for the convergence of data and storage networks onto a single lossless Ethernet network. Integrating the application virtualization manage-ment with the network enables the ability to move virtual machines across the WAN providing for disaster recovery and load balancing between data center sites.

Storage

Data Center HypervisorRock Mount

ServersBlade

Servers

Data CenterInter-Connect

Data CenterSwitching Core - OS10 OSK

TOR - OS6900(Mesh)

ServersStorage

WAN[DWDM 1830]

Service Routers[77x0 SR]Photonic Switch [1830]

FCoEiSCSI

Hyper-V Server

= =

Management

Visibility[VitalSuite]

Service Aware Mgr

Network Management[OmniVista]

Address Mgt[Vital QIP]

Authenication, Authorization& Accouting[8950]

Virtual Machine Mgr[OmniVista 2500]


11

managed Communication service networksMore and more small enterprises are looking to service providers for a complete all-in-one com-munications solution that includes voice, data (wired and wireless), collaboration tools, and video. To meet these customer expectations, service providers must provide:

• Easyplug-n-playdeploymenttoreducecustomeracquisitioncost

• ComprehensiveremotemanagementcapabilitiestoreduceOPEX

• RichQualityofExperience(QoE)toensurevalueandcustomersatisfaction

Figure 9: Managed Communication Service Networks

Figure 9 shows a typical managed communication services network based on the application fluent network. Such network varies from the hosted application platforms to the end points at the cus-tomer’s site and also includes the complete LAN environment normally behind the NAT and firewall.

aPPliCation fluenCy: tHe next generation of HigH leverage network infrastruCtureThe demanding requirements in the business networks, distributed data center, cloud, rich media applications, and mission-critical applications combined with the limitations of legacy networks are forcing operators to look for innovative ways to improve network efficiency and performance while minimizing operating costs and capital investment. The application fluent network as defined under Alcatel-Lucent’s High Leverage Network vision promises to transform the way operators enable different applications over a single infrastructure using the OmniSwitch portfolio, to reduce com-plexity while increasing the overall efficiency. The OmniSwitch, with application fluent architecture, features a resilient architecture, streamlined operations and automatic control, which together meet the challenges facing operators today. With tangible benefits such as flat architecture, simplified ordering options, reduced demands on network engineers, rich QoE, and a faster time to ROI, the OmniSwitch powered application fluent network represents market-leading innovation in network design and operation — one that Alcatel-Lucent is uniquely qualified to lead.

PC w/ SIP Client

IP Phones Ethernet WAN

xDSL

Local Call*

SMB

ISP OA&M

ISP

Wi-fi phones

SIP Servers

ISDN Phones

SMB

ISDN BRI

ISDN BR/PRI(E1 or T1)*

ISDN BRI/PRI(E1 or T1)*

TR-069/Motive

SCM, ACS, AAA, NTP,SSH/HTTPS client

LANOmniSwitch/OmniAccess Alcatel-Lucent SIP servers,

BroadSoft,....

PSTN

TDM PBX/IP PBX

SCM: Server Configuration ManagerAAA: Authentication, Authorization & AccountingACS: Auto Configuration ServerNAS: Network Access Server

www.alcatel-lucent.com/ Alcatel, lucent, Alcatel-lucent, and the Alcatel-lucent enterprise logo are trademarks of Alcatel-lucent. All other trademarks are the property of their respective owners. the information presented is subject to change without notice. Alcatel-lucent assumes no responsibility for inaccuracies contained herein. copyright © 2011 Alcatel-lucent. All rights reserved. 2011123754 (December)

wHy alCatel-luCent omniswitCH Portfolio? Alcatel-Lucent is taking the lead in the industry’s migration to the application fluent network. Why is Alcatel-Lucent’s OmniSwitchTM portfolio the best choice for building the application fluent network?

• Alcatel-Lucent’sOmniSwitchproductlineisanaward-winningcomprehensiveportfoliobuiltonasingle operating environment which addresses operators varied infrastructure needs over Ethernet.

• Alcatel-Lucent’ssystematicmultivendorapproachmaximizesexistinginvestments.

• Unlikethecompetition,weofferdisaggregatedsolutionsthatprovidechoicesandallowoperatorsthe ability to tailor the network architecture to meet their specific needs

• Ourcommitmenttoanopennetworkinfrastructureenablesamultivendorstrategyby:

o Providing rich interfaces into the OmniSwitch portfolio

o Supporting industry standards

o Accommodating third-party management tools

• SinglepaneofmanagementacrosstheOmniSwitchandAlcatel-Lucent’sopticalandservicerouterportfolio to enable end-to-end provisioning and QoS.

• Ourprofessionalservicesareavailabletoassistwithspecializedknowledgewhenneeded.Ourconsultants are trained in a wide range of technologies and platforms, allowing them to effectively support complex, multivendor network installations.

• Ourexpandingecosystemaugmentsthevaluedeliveredtoourcustomer.

building high leverage networks

Documents