How To Choose a Frame Relay / ATM Provider

It is widely understood that a single vendor strategy brings corporate data communications organizations numerous benefits, including improved interoperability and security, better reliability and availability, ease of introducing new services and technologies, and lower cost of ownership and operation. Many of these advantages are extremely hard to realize if the network managers are trying to stitch together the point products of numerous competing vendors. A single vendor that can provide all the solutions and technologies solves this problem and can be an essential part of a corporation's strategic plans.

When a corporation selects a service provider whose network is also based on the same vendor's equipment, the advantages can multiply still further. This white paper discusses the changing demands on corporate networks, explains how a single-vendor service provider network differs from those based on point solutions, provides examples of benefits of a true end-to-end approach, and shows how to find the best service provider for you. It also describes the key roles that Cisco plays in most of the large telephone companies and service providers around the world, and the benefits you can gain from using the services of a Cisco Powered Network.

Changing Times, Changing Requirements

Traditionally, corporate wide area networks have been built with leased lines interconnecting customer-owned equipment at user locations. Protocols such as X.25 and Systems Network Architecture (SNA) were used in building networks that spanned the world and could handle mission-critical traffic for the enterprise.

More recently, the emergence of fast packet technologies such as Frame Relay and Asynchronous Transfer Mode (ATM), plus the technologies of the Internet (such as TCP/IP and HTML), have brought significant changes to this pattern. This technical change is taking place at a time when user application bandwidth requirements are escalating. Old network technologies are rapidly being superseded, but managing the transition is not a simple task. Budgets aren't growing as fast as user demands, so that means effective management of resources---including tailoring networks for different qualities of service (QoS)---is essential. For information systems and communications professionals, this means a major set of challenges, as well as opportunities.

As part of this changing environment, large corporations are increasingly considering outsourcing their voice and data networking requirements to service providers. Of course outsourcing is a growing trend in many areas, but in net- working, this movement is driven by increasing demands from user departments, by operations costs and staffing shortages, and by the problems of keeping up with the fast pace of technology evolution.

Top Ten Reasons Why Corporations Outsource:
1. Improve company focus 2. Gain access to world-class capabilities 3. Accelerate re-engineering benefits 4. Share risks 5. Redirect scarce corporate resources 6. Conserve capital 7. Gain cash from sale of assets 8. Reduce and control operating costs 9. Avoid recruiting and retaining specialized staff 10. Eliminate management of difficult / complex functions Source: The Outsourcing Institute

As they consider options with service providers, customers look for reliable services that have the "look and feel" of a private network, are easily adaptable to their business needs, and are delivered to them at a reasonable price. Operations and engineering staffs are today finding themselves increasingly involved in working with service providers. This trend toward increased utilization of network services, or even complete outsourcing of the network, is driven by several key trends. Networks that perhaps once mainly linked batch computer operations are now supporting real-time, interactive, and increasingly mission-critical applications. The continued proliferation of PCs, LANs, and intranets is steadily decentralizing the installation of networking equip- ment, with the attendant problems of administration and management. And it's getting harder and harder to hire, train, and retain qualified staffs of network technicians and engineers.

The role of engineering and operations groups in achieving their companies' business objectives is as important as having good services and effective marketing. Operations and engineering groups are asked to build and operate networks that:

• Are stable and manageable, yet can be easily upgraded

• Can scale to meet increased user and bandwidth demands

• Deliver extremely high reliability and availability

• Seamlessly integrate new technologies along with the legacy equipment

• Can quickly and easily accommodate new services

• Flexibly handle multimedia and other quality-of-service requirements

• All while managing resources to "do more with less"

In addition to issues of security and legacy network migration, the problems of skill shortages among staff and dealing with complex technologies are key reasons why network outsourcing is becoming increasingly popular. A recent Yankee Group survey of 100 multinational corporations found almost one-third reporting these as major problems.

Of course there is hesitation about increased outsourcing among corporate network managers as well, because while the benefits of outsourcing their networks may be obvious, so are the concerns that visibility and control of their networks will diminish and quality of service to their users will suffer.

From a security perspective, corporate network management sets policies for access, use of applications and network resources, and data security. When corporations outsource their WAN requirements to service providers, the service providers not only provide some of the resources, but must in essence enforce the customers' policies as well. Since security encompasses not only access, but data privacy, assurance, and integrity, it must be implemented end to end, including both the network and the customer premises equipment (CPE). Of course security policies need to be consistently implemented, regardless of the way a user accesses the network, both in the service provider's network and in the user's equipment.

A service provider that can provide a comprehensive set of security services gives customers confidence in the network and in their service offerings. But achieving this is extremely hard to do if the service provider is trying to assemble its network from numerous vendors' one-off products.

Users want to define quality-of-service policies that can be implemented both in their CPE and in the service provider's network. Cisco helps service providers do this with the most complete set of traffic management tools available. These fit user needs, since user applications themselves can seldom request or require strict QoS by themselves. And the same capabilities allow service providers to better scale their networks.

The times are changing for service providers as well. Once mainly concerned with plain old telephone service and a few leased lines, they were practically invisible within the corpo- ration, with their influence extending no further than the CSU/DSU. Today, Frame Relay, ATM, Internet Fax, Managed Router Services, and similar sophisticated services bring service providers ever further into the enterprise.

Service providers urgently need to retain large, high- value customers, and in fact expand into new markets. The telecommunications industry environment has become very competitive, with service providers offering their corporate customers tailored, highly reliable services, along with a lower total cost of ownership. Innovation and service-level agreements are the order of the day with, for example, a major carrier posting its virtual private network (VPN) performance information on the World Wide Web, and offering a 50 percent rebate of the month's charges if a customer experiences an outage of one minute or more during the month on its dedicated access services.

Frame Relay initially became popular replacing point-to-point leased lines. It provided a flexible and efficient data transport solution while simultaneously reducing the cost of the overall network. The replacement of SNA and other legacy networks, while usually undertaken as a cost initiative, brings other benefits as well, in that a fully meshed network (with greater reliability) often becomes practical. The emerging trend is to use Frame Relay networks to multiplex both SNA and other protocols over Frame Relay virtual circuits to achieve even greater network efficiency. Such network changes don't happen overnight, and usually require a phased, and well-coordinated, approach over many months.

But all of these changes also require the corporate networking staffs and the service providers to work together more closely than ever. There are numerous examples of this, including the mapping of SNA classes of service onto network QoS parameters, the interpretation of Customer Network Management (CNM) service information by customer-owned network management systems tuned to the Cisco-based premises network, the operation and administration of seamless multicast capabilities across the LAN and the WAN, and the interoperation of premises-based directories and network-wide directories operated by the service provider. On the electronic commerce front, service provider-provided certificate authorities must seamlessly interwork with premises network IP security (IPsec) implementations. When corporate networks handle integrated voice/data/video services, interworking of premises-based multi-service access equipment with integrated network services becomes critical.

In the area of voice-over-Frame Relay, careful planning and coordination between the different parts of a network is essential. Quoting the Frame Relay Forum: "In the absence of firm definitions for vendor equipment interoperability, or transport of voice across a carrier's public Frame Relay network, many equipment vendors have developed proprietary methods for integrating voice onto Frame Relay networks. Each vendor has, in their products, had to address issues such as compression, echo cancellation, delay and delay variation, frame loss, and traffic prioritization, often through different hardware and software implementations."

Once upon a time, the telecom industry was highly compartmentalized. Telephone companies had analog voice networks, cable TV operators had analog television networks, and Value Added Networks (VANs) offered hybrid analog/digital data services. Every network had its own protocols and was pretty much independent. It is increasingly clear, however, that packet techniques and Internet-based technologies (for example, Frame Relay, ATM, and TCP/IP) will be the primary basis for services in all types of networks, both private and public.

Aren't All Service Providers Pretty Much The Same?

When selecting a service provider, one often thinks of three key questions: does it offer the service we need, does it offer it in all the locations we require, and does it offer it at a fair price? But there actually are additional considerations that can greatly impact both your cost and performance results.

A corporation can gain many benefits from carefully selecting its service providers. A recent article in Network Magazine (www.network-mag.com) notes that the most important criteria should be the service provider's network infrastructure. If a service provider has chosen to build its network using a single-vendor strategy, it gains improved interoperability and security, better network availability and reliability, ease of introducing new services and technologies, and lower cost of ownership and operation.

When a company looks for a service provider for Frame Relay or ATM services, the goal is for services that are both cost effective and essentially transparent to users. A service provider that has based its network on Cisco equipment offers just that. After all, corporations with data commun- ications networks probably already know a lot about Cisco Systems and its many products and solutions. It is easier to take advantage of new service provider offerings and features when that service provider has chosen to base its network on Cisco Systems.

Cisco goes to great lengths in designing every element of a service provider's network to consider 24 by 7 reliability and operations concerns. Many Cisco products are available in configurations that meet Bellcore's "NEBS requirements," with redundant power supplies and other elements. Onboard hardware diagnostics and LEDs show status, utilization, and service conditions at a glance, for easy visual troubleshooting. Flash memories are used to allow remote downloads of software without hardware changes, thus reducing costs of administration. Plug-in modules can be "hot swapped" as required, helping maintain service. On many systems, even software upgrades don't interrupt service.

It turns out that all service providers are not the same. Even when services, locations, and prices are comparable, a service provider whose network is built with a single vendor gains many benefits, and these can directly favor customers. Here are a few examples.
Table 1: A Single-Vendor Network Infrastructure Strategy

Factor	Service Provider's Benefit	Which Gives You
Performance	Consistent high network throughput	Fast access/ operation of services
Interoperability	Minimal system integration problems	Fast response to service requests
Security	End-to-end functionality	Secure and private data
Availability	Less network downtime	Round-the-clock traveler access
Reliability	Faster restoral of network failures	Support for mission-critical users
Support	Single vendor responsibility	No "finger pointing"
New Services	Competitive positioning	Best choices in the marketplace
Technology	Easier migration of network and services	Innovative services with minimal interruptions
Training	Greater technical proficiency	Effective technical support
Lower Costs	Less investment and expense per user	Most competitive prices

Naturally you'll want to establish Service-Level Agreements (SLAs) with your chosen service provider. The following are some common SLA topics. The key is that a single-vendor strategy means these service-level agreements are more likely to be realized.

SLA Topics

Installation (when? trial period? how coordinated with previous provider?) Responsibility limits (demarcation points? includes software issues? intercarrier issues?) Service availability (24 hours a day? 5 days a week? 99.99 percent of the time?) Service performance (in terms such as bandwidth, delay, etc.) Service response time (24-hour help desk? 4 hours to technician on site? escalation?) Service-level agreements (written? penalties? how measured or verified?) Interfaces (standards-based or proprietary?) Upgrades (frequency? when during day/week implemented? advanced notice? cost?) Disaster recovery (their network? your site? loaner equipment pool?) CPE supported (one vendor only or several? trade-in of existing?) Billing (how often? what increments? volume discounts? refunds for failures?) Network management (monthly reports? Dial access to SNMP data? Web-based access?) Network / service design (consultation? periodic reviews? costs?) Training for your staff (initial? follow-up? with upgrades?) Contract cancellation (term? if poor service experienced?)

Recognizing the challenges for its corporate customers in selecting the best service provider, Cisco has launched its Cisco Powered Network program. This program enables qualifying service providers to use a special Cisco Powered Network logo, shown here, in their advertising, on their Web site, and so forth. When you see a service provider displaying this special logo, you'll know that its network is based on Cisco equipment, and you can be sure of its strong capabilities in all of the areas described above, as well as in the innovative services it has available.

 

How a Cisco Powered Network Is Different

If you are considering two service providers who offer the service you need, at the locations required, and at comparable prices, why would choosing the one displaying the Cisco Powered Network logo give you significant advantage? How does the service provider achieve better reliability and security, more innovative services, and so forth? We will describe how a Cisco Powered Network is superior from several perspectives, beginning with how service providers are evolving their network architectures, then looking at the network edge, the network backbone, and services-aware management, and concluding with examples of interaction between your network and that of your chosen service provider.

Evolving Network Architectures

To become the network of choice, service providers around the world are moving toward a future network architecture that accommodates circuit-switched and packet-switched elements and appropriate gateways between them. This "network of the future" features a clear distinction of functions and roles between the edge and backbone elements, as well as a unifying directory and well-defined control planes.

At the edge of the network, customer access is aggregated, service policies (including security and quality of service) are defined and applied, and data collected for both billing and engineering purposes.

In the backbone, fiber, Wavelength Division Multiplexing (WDM), and SONET/SDH will be the norm, with a focus on high performance, management of traffic flows (consistent with both quality of service and type of media), and congestion management.

The directory for the network offers an integrated way to identify all network users and devices. Cisco is extending this traditional directory concept to encompass policy, application, media, protocol, and six new classes that define how a user interacts with the network. This future network architecture is appropriately referred to as a "directory enabled network."

The importance of software in linking the different elements of the network together cannot be overemphasized. Millions of decisions are made second by second in a service provider's network, and it is the software making these decisions, regardless of the type of hardware or device involved.

At the Network's Edge

Of course there are many operations and administrative tasks made simpler with a Cisco Powered Network. For Frame Relay service, Cisco's auto-install capability automates operations and allows routers to be quickly and easily installed from a central management location. Cisco support for RFC 1293 (Inverse ARP over Frame Relay) allows the router to automatically discover Data-Link Connection Identifier (DLCI) addresses and to build its own routing tables based on the exchange of information with other RFC 1293-capable routers. Furthermore, software upgrades can be broadcast easily to all nodes on the network in off hours.

Collectively, these features, standard with Cisco but often not found in generic Frame Relay access devices (FRADs), substantially reduce the amount of research, preparation, and manual intervention required for the service provider to bring a new unit on line. The larger the network, of course, the greater the benefits for operations, and the faster customer needs are met.

Cisco's ability to auto-install, auto-configure, automatically discover topology changes, and automatically distribute soft- ware upgrades can save a typical operations budget easily two to three hours per unit per year.

As the leading networking vendor and the supplier of 80 percent of the Internet's routers, the Cisco configuration methodology is widely understood. For a service provider, training technical staff or finding additional staff familiar with, and/or interested in learning, Cisco configuration, is relatively simple. On the other hand, the configuration methodology employed in many generic FRADs can be both complex and time-consuming. In fact, one leading FRAD is so difficult to configure it reportedly challenges the company's own field reps.

A key factor in reliability of a service provider's network is the robustness of the hardware platforms used. Cisco FRADs and access routers are based on one of the most widely used, mission-proven platforms available today. Close to a million units have been installed throughout the world, all based on stable designs and millions of years of experience. In contrast, many FRAD vendors' products are based on off-the- shelf PC motherboards and components, not noted for reliability, and often varying from month to month from different suppliers.

Cisco IOS software, a de facto standard worldwide, is another key element in service provider network reliability. Unlike its imitators, the Cisco IOS software has been thoroughly proven in mission-critical environments and across a comprehensive range of networking applications. No vendor supplies a wider range of software features and capabilities.

Network Backbones

A hallmark of a Cisco Powered Network is that it doesn't depend upon just one technology or architecture. Choices for service providers include a pure router backbone (often used by ISPs), or an IP-over-ATM backbone (where Cisco's Tag Switching offers significant performance advantages), or a multiservice backbone that simultaneously provides IP, Frame Relay, and native ATM services. While the long-term trend is toward multiservice networks, managing multiple networks (and their interworking) will be required for many years.

One of the key requirements in the backbone is congestion management. Layer 2 switches, such as the Cisco BPX^? 8600 Series wide-area switch or Cisco LightStream^? 1010, have sophisticated congestion management capabilities for both Frame Relay and ATM services.

For backbone networks using ATM, managing virtual circuits and bandwidth allocation manually is difficult and time consuming. Cisco offers its Automatic Routing Management software, which completely automates these functions, reducing network operating costs and improving network reliability. This software is perfect for service providers---it is highly robust with no single point of failure and is distributed throughout every switch in the network.

Customer complaints are at a minimum with Cisco's Optimized Bandwidth Management congestion avoidance and bandwidth optimization technique, based on a full virtual source/virtual destination implementation of the ATM Forum's Traffic Management 4.0 specification. It is a cell- based, closed-loop feedback mechanism that continuously monitors the utilization of interswitch trunks and dynamically adjusts the bandwidth of all virtual connections, completely transparent to users. In addition to improving network utilization it improves the reliability of variable bit rate (VBR) data services (for example, bursting capabilities) by mini- mizing retransmissions caused by cell discard. Accomplishing both goals at once, as Cisco's systems do, is impossible with switches from most vendors.

Of course in the backbone, as well as at the edge, scalability is crucial. All Cisco systems are highly modular, achieving maximum flexibility in designing networks to meet evolving service requirements.

And reliability is enhanced by Cisco features such as hardware redundancy features, hot insertion of plug-in modules, and the ability to download software upgrades and fixes.

A last thought about network backbones addresses typical SONET/SDH carrier transmission facilities. Cisco believes that future networks will be built primarily upon fiber facilities, often with Wavelength Division Multiplexing (WDM) used to gain additional bandwidth. SONET/SDH terminals will support some higher-layer networks, including the circuit-switched public telephone network, some ATM networks, and some IP traffic directly mapped to the SONET/SDH payload. Cisco has solutions in all of these areas now, and is pursuing active development programs for each of them. With a Cisco Powered Network, corporate network managers and service providers have the opportunity to implement the ideally designed network for their particular circumstances, and they know that Cisco can properly support it.

Services-Aware Management

The International Telecommunications Union (ITU) is the main standards body for service providers worldwide. The ITU recommends a multilayer approach to network and services management, ranging from element management (for example, alarms) at the bottom to business management and billing at the top. Collectively these functions are called the Telecommunications Management Network, or TMN.

Most corporate data networks, on the other hand, rely on Simple Network Management Protocol (SNMP)-based network management systems, and corporate voice networks use yet other management techniques. And of course different vendors often have their own ways of implementing each of these standards. Cisco has recognized the need of service providers to not only manage their own networks, but to have the capabilities to offer customized network management capabilities to their corporate customers. In response, the Services Operations Management System (SOMS) has been created to guide developments both within Cisco and in Cisco's partners. The following are the kinds of questions this organizing framework seeks to answer.

Key Issues of Operations Management

How do I tie customers/users to underlying services provisioned in the network? How do I manage and assure these services end-to-end across a diverse set of networking technologies? How can I quickly and efficiently deploy new services across a diverse network infrastructure? How do I manage growth and change in this infrastructure? How do I maintain the integrity of the service infrastructure? How do I account for and bill for usage? What is the "glue" around which these applications will interoperate and around which I can build and extend their value? How do I bring new services to market more rapidly, helping my company to increased profitability? How can I offer services tailored to individuals as well as to groups?

One example of Cisco's answers to these needs is the Netsys Service-Level Management products, focused at the services management level of the management hierarchy. Netsys is the only product offering on the market that maintains a compre- hensive understanding of the total network configuration, packet flows, and utilization across routers and switches. This workstation-based tool helps service providers establish end-to-end network policies, ensure connectivity between key network nodes, provide appropriate security and protection for network resources, and confirm overall network performance at both the network and application level. From the operations perspective, it provides integrity checks, path security analysis, specific performance analysis, and guided repair, based on a network model and a reasoning engine driven by periodic uploads of actual network configuration and statistics data.

The service provider establishes policies with its customers that explicitly define the levels of availability, performance, and security necessary for each service and application. The Netsys software assesses these service requirements automatically on a scheduled basis established locally. The resulting Web-based report card quantitatively summarizes which requirements are being met and which have been violated. Violations can be quickly sorted by severity, type, and other criteria. An integrated, end-to-end view of the network---showing both Layer 2 and Layer 3--- makes it easy to spot areas impacted by service-level violations.

The service provider's operations staff can then easily drill down to isolate the problem and use the built-in QuickSolver capabilities to suggest the most effective corrective action. An offline network simulation engine, precisely modeling the service provider's actual network, allows safe offline testing of proposed changes, upgrades, and so forth.

As another example, Cisco Voice Manager is a Java-based application that simplifies the process of deploying and managing end-to-end, voice-enabled networks. It works with the Cisco 2600 or 3600 series routers to provide unique capabilities for monitoring and managing end-to-end multi- service networks. Cisco Voice Manager takes the complexity out of large voice-over-IP networks by simplifying the process of configuring voice and fax interfaces and administering the voice dial plan. Cisco Voice Manager provides Call Detail Records, Call Volume Reports, and Active Call Reports, and measures QoS parameters such as delay, packet loss, and type of service.

Another critical aspect to operations and engineering personnel is Cisco's move toward a Directory Enabled Network (DEN) architecture. Working with Microsoft and other firms, Cisco is creating a flexible and scalable directory service for use in managing network infrastructures, providing richer network services, and easing operations tasks. For service providers, this means lower total cost of ownership, single points of administration for all network resources, and the ability to cost-effectively provide differentiated network services with individualized policy control. The DEN concept originated by Cisco is now supported by most major vendors.

How can a service provider deliver the services that corporate users increasingly demand? Many service providers have discovered that working with a single strategic vendor can significantly enhance their ability to do this. They have found such a strategy helps them deliver high-quality service while keeping their costs---and thus the customers' prices--- competitively low. Cisco has launched its Cisco Powered Network program to both strengthen how it works with service providers in this way and to promote the services of those service providers who qualify. The Cisco Powered Network designation is available to service providers who maintain and publish high standards of service and whose services rely mainly on Cisco equipment. The large number of service providers who already display the Cisco Powered Network logo know, and show their customers, that they are among the best of the telecommunications world.

Going End-to-End with Cisco

Most companies have already discovered the advantages of relying on Cisco equipment, solutions, and services to meet their networking challenges within their organizations. A company that has built its internal networks around Cisco products, and in addition, uses a Cisco Powered Network service provider, gets the best of all worlds. This is true because in developing products and solutions for both enterprises and service providers, Cisco is focused on the added value that can be provided in end-to-end solutions.

Cisco development goals include:

• End-to-end availability---Nonstop core networks, end-to-end network testing, diagnostic tools, network backup strategies, and robust protocols. Coupled with Cisco's support services. the network downtime and the difficulty of resolving issues can be minimized.

• End-to-end QoS---Per virtual-circuit-queuing, closed-loop ABR support, mapping of IP and Frame Relay/ATM QoS, enhanced LMI protocols, and congestion management response using WFQ and WRED. The complexity of delivering packets end to end with the QoS required is high and requires an end-to-end architecture. Coupled with Cisco's consulting, end-to-end QoS is achievable.

• Simplified end-to-end management---Consistent element management, integrated alarm management, and integrated provisioning. Cisco is taking on the industry challenge to rationalize and integrate management systems and will be the first to deliver an end-to-end view of the network.

Cisco can bring unequaled capabilities to help engineering and operations departments with the products and tools to achieve their goals. The purchasing process is faster (much can be done on the Cisco Web site, for example), equipment is installed and put into service more quickly, and integration testing is vastly simplified. A Cisco network solves issues of interoperability, reduces complexity, and allows significant cost savings on training, testing, trouble- shooting, and support. (Studies repeatedly show that these factors, not initial cost, have the greatest effect on total cost of ownership.)

Cisco's full range of products and services stress the importance of manageability, security, and resource controls; address the entire network, including access, distribution, and backbone; and support effective resource management, including reliable and accurate billing, allowing corporations to charge back network costs to user departments. Cisco provides leading-edge technology, more than a decade of Internet experience, and worldwide design, consulting, and training support.

The advantages of the end-to-end approach include easier adoption of new technologies and services, a more secure network, lower operations and maintenance costs, better overall network performance, and end-to-end service assurance. We'll address each of these areas in turn.

Companies already know how Cisco protects their investment with highly scalable, high-reliability products, integrated with the Cisco IOS software. Increasingly, the offerings of service providers will be software-based, and having IOS software at your end as well as in the service provider's network makes for much faster implementation of new features. Cisco IOS software is a de facto standard, and new features will be operable in IOS software before other vendors catch up with the standards. And when you connect your Cisco-based corporate network to the Internet, no special client software is required, greatly easing administrative and support burdens.

Implementing consistent security and other network policies across your own facilities plus those of one or more service providers isn't an easy task. But when all of the network elements are based on Cisco IOS software, many of the potential headaches are minimized.

Looking to lower your networking costs? A Cisco Powered Network service provider not only increases network efficiency, but virtually eliminates finger-pointing between your staff and the service provider, with Cisco engineers able to help solve problems regardless of what---and where---they are. Common terminology goes a long way toward eliminating misunderstandings on service orders and trouble tickets, and your team spends less time in training classes, studying manuals and documentation, and so forth. Network managers know that installing heterogeneous point solutions for specific facets of their networks backs them into a corner from an operations perspective. The expertise and experience of operations and engineering staff is a critical resource, but if it cannot be flexibly applied to operating and maintaining the entire network, efficiency of the whole organization suffers. Analysts have referred to this situation as the "Balkanization" of network management---where skills become so compartmentalized that each technician becomes responsible for a specific set of equipment. Every time a new type of equipment is added, the learning curve starts all over again.

One dramatic illustration of the advantages of a single vendor at the corporation and in service provider networks is the Year 2000 problem. While a lot of attention to date has been on computer systems themselves and their applications, many of the same issues are present to some degree in networking equipment and systems. Operations and engineering effort for projects like Year 2000 are greatly reduced with fewer vendors' products to investigate, sequence upgrades, and so forth.

With common technologies and interfaces, on the other hand, these operations problems are avoided. The network can be more effectively managed, and service levels and reliability increase. At a time when hiring and retaining competent technical staff is a challenge, the ability to amortize the skillsets of individuals over the maximum range of network equipment and services is a compelling reason to apply consistent technology throughout the network.

And all of these operations advantages apply equally well when a corporation is working with a service provider on its overall network.

The performance of your network is enhanced as well, especially when you consider the growing range of user applications. Uniform implementation of features such as quality of service and voice/data integration between your equipment and that of the service provider means less integration effort and faster project implementation. From a technology perspective, common software and assured interoperability means much less hassle, time, and risk in trying to integrate the products from various vendors as the network grows.

And how do you know that you're getting what you're paying for? The Cisco Netsys service-level assurance tools described earlier enable you to have an end-to-end view, with visibility at both Layer 2 and Layer 3, of your entire network, including the service provider portion.

Conclusions

Operations and engineering personnel play a key role in the business success of any corporation, and minimizing the number of network equipment vendors is a powerful strategy in achieving these goals.

When a Cisco Powered Network service provider is added to the mix, the benefits grow even more. This program illustrates Cisco's commitment to truly take the end-to-end view, including the service provider in the middle of your network. This means, for example, that Cisco will:

• Incorporate an end-to-end services perspective in all developments, both for enterprise and campus products as well as in those designed for service provider networks

• Offer both single- and multiservice access products for Frame Relay, ATM, and IP services

• Provide infrastructure-class products that are flexible and support open standards

• Provide a rich, robust set of service enablers such as security, VPNs, and so forth

• Support IP over multiple technologies such as ATM, SONET, and WDM

• Provide world-class consulting and support globally

You need to choose your service provider wisely, and you need a strong technology partner who understands your business needs and can provide solutions to the functional and operational challenges you face. Cisco has an unmatched record of leadership and expertise in building robust, high-reliability networks for both the world's largest private networks and for service providers worldwide. And Cisco brings partnerships as well, ranging from global strategic relationships with Microsoft and Intel to those that are more tactical and implementation focused ones such as with HP, EDS, and KPMG.