Today, voice, data, video, and interactive collaboration applications traverse the same network infrastructure and extend beyond the Local Area Network (LAN), Wide Area Network (WAN), and extranet. This diversified traffic has greatly increased the role and importance of applications management within networks. Legacy point-to-point network implementations designed for LANs are giving way to real-time, rich media, interactive, and globally deployed business applications that can be accessed by an organization's employees, partners, suppliers, and customers. These applications must be dependable, secure, and offer ubiquitous access to support critical business processes.
This white paper presents key market trends that point to the need for end-to-end application performance management within service provider networks, which a 2009 Nemertes Research issue paper defined as the ability of a network "to relate application traffic to business processes and offer enhancements or services in support" of those business processes. The Cisco® Application Performance Management Framework provides application performance management through an array of capabilities within existing Cisco and partner platforms, technologies, solutions, and the Cisco IP Next-Generation Network (IP NGN) architecture. It provides a significant competitive advantage for managed service providers as they help customers to deploy highly-scalable applications linked to mission-critical business processes with stringent Service-Level Agreements (SLAs) in a more efficient, flexible, and secure manner.
Market Overview
Businesses of all kinds recognize the productivity benefits of real-time, collaborative, and secure applications. They want these applications to be accessible to an ecosystem of employees, partners, and customers, whether in the office, at remote locations, or on-the-go via mobile devices. This imperative is heightened in the current economic downturn, with the drive for more efficiency from employees and existing assets, because it may be vital to a business's survival.
Cisco has identified three important trends that will affect enterprise IT spending in the next few years and drive the demand for intelligent, application performance management networks:
• Consolidation of IT Infrastructure and Centralization of Applications, security, and processes into data centers. This consolidation will allow managed service providers to offer unified network fabric and integrated provisioning services tailored for enterprise data centers.
• Virtualization of Services in the Network Cloud, driven, in part, by the increasing prevalence of Service-Oriented Architectures (SOA) that will allow managed service providers to offer IT-as-a-Service (ITaaS), Software-as-a-Service (SaaS), and other virtual services that are heavily reliant on scalable, optimized, and secure applications delivery.
• The Borderless Enterprise, which is being driven by the permeation of Web 2.0 technologies and the requirements of the empowered user/knowledge worker. Empowered workers desire instantaneous collaboration at work, home, or on the go, and greater control over the applications they use.
All of these trends are being driven by the need to trim costs while adding new services and efficiencies that promote business competitiveness. All these trends necessitate major changes in how applications are deployed and managed on networks.
As applications have moved beyond the LAN to the WAN and the myriad of end-user devices, there is a greater reliance on the network to deliver them efficiently with Quality of Service (QoS), optimization, control, security, and management. Business applications such as email, Customer Relationship Management (CRM), and Enterprise Resource Planning (ERP) systems, and other bandwidth-intensive applications such as streaming video, telephony, and point of sale or web-based transactions all have different traffic characteristics and levels of priority. Most enterprise applications were originally designed for the LAN, where the users and application servers are in proximity. In this environment, issues of distance and access to applications and resources across closed-user groups were not relevant.
In the migration of services from the LAN to the WAN for the borderless enterprise, other challenges include the following:
• Data Center Consolidation was designed to reduce total cost of ownership by reducing the equipment at branch offices. These consolidated resources must provide higher levels of security to comply with industry regulations (such as Sarbanes Oxley and the Healthcare Insurance Portability and Accountability Act or HIPAA) and must be optimized to make the most efficient use of bandwidth. Certain meshed application flows (e.g., peer-to-peer, voice, and video) may bypass the data center and require monitoring and performance management to achieve LAN-like response times.
• Service-Oriented Architectures represent a componentization of the middleware layer of an application to provide broader connectivity to the database and other applications, shifting application delivery from the server to the network. This new architecture challenges the WAN to support end-to-end performance more effectively.
• Virtualization of servers and utility or cloud computing architectures require the network to become more dynamic and agile to support capacity planning, dependable performance, scalability, security, and application QoS.
According to a 2008 study by Forrester Research, data center managed services are growing dramatically and projected to reach US$96.7 billion in revenue worldwide by 2013, a compound annual growth rate (CAGR) of 17%, Figure 1.
Figure 1. Projected Growth of Data Center Services: Main Categories
Source: Forrester, 2009.
Among individual data center managed services, content and application acceleration show the greatest growth through 2013, with a combined CAGR of 20.9%. Within this sub-segment application acceleration services (for applications such as Enterprise Resource Planning [ERP], Customer Relationship Management [CRM], and supply chain) show the largest growth, with a CAGR of 25.9% by 2013, Figure 2, with content acceleration services (for web pages with rich media such as audio and video) projected to grow at a CAGR of 18.7% by 2013.
Figure 2. Projected Growth of Data Center Application and Content Acceleration Services
Source: Forrester, 2009.
In a 2008 study by Chadwick Martin Bailey research, enterprise customers were asked if they would consider managed and hosted services for greater application control and efficiency. A high percentage of respondents said they were extremely likely to do so, Figure 3. The results also showed that the smaller the organization, the greater the interest there was in application services.
Figure 3. Adoption Plans for Application Technologies
Source: Chadwick Martin Bailey study, 2008.
A New Framework for Application Performance Management
The concerns over application performance and trends such as data center consolidation, globalization of information and communications resources, virtualization, and new models for service delivery mentioned earlier collectively demand a new framework for managing and helping ensure application visibility, performance, control, and security across the network. These concerns require service providers to have an application performance management network with the ability for the total network infrastructure (platforms, applications, switches, storage, capacity, and mobility features) to operate in an integrated manner to meet these requirements.
Application performance management provides a framework to address some of the requirements for application performance management. A principal mechanism for delivering application performance management is application-aware networking. It relies on a combination of technologies, capabilities, and services in which every device on the network has the intelligence to treat discrete applications appropriately end-to-end. Application performance management provides the basis for comprehensive solutions that enable application performance management within networks, making today's networks and applications more flexible, reliable, efficient, secure, and scalable.
The Cisco Application Performance Management Framework integrates network and application intelligence and the use of application delivery networking techniques. Gartner defines application delivery networking as the use of WAN optimization techniques (using WAN optimization controllers) and application delivery control techniques (using application delivery controllers) to help ensure the delivery of application-based services end-to-end in a secure manner.
Among other methods and processes, these techniques include:
• TCP Multiplexing a connection-pooling method that reduces the overhead required to establish and tear down TCP connections with servers, improving application responsiveness.
• TCP Optimization mechanisms for improving TCP performance.
• Data Compression and Caching HTTP compression, proprietary data reduction algorithms, and caching help optimize the flow of application data.
Cisco has integrated many other features in the Cisco Application Management Framework, including the use of business policies to enforce requirements for the end-to-end management of applications. The framework spans end-user devices wherever they are, as well as provider edge, network core, and data center environments.
The functional hierarchy of the framework includes three tiers, Figure 4.
Figure 4. Application Performance Management Framework - Visibility, Optimization and Control Services
Application Visibility solutions provide real-time performance monitoring to discover applications running on the network, to understand how those applications utilize network resources, and to measure the performance of critical applications.
Application Optimization capabilities include compression, caching, TCP flow optimization, and application optimization techniques to enhance application performance and optimize WAN resource usage. These solutions can significantly improve the performance of applications in branches, centralized data centers, and application access by mobile and remote users. Application acceleration and optimization enable data center consolidation, which offers simplified management and support, and adhere to centralized compliance models.
Application Control gives service providers dynamic and adaptive tools to monitor and help ensure application performance. Critical applications can be prioritized, and non-business-related applications can be treated with lower priority. Special policies may be dynamically provisioned for customers, and the security behind the firewall may be extended to help ensure protection of information resources. Application-aware path optimization matches critical applications to the best path, so that application performance and availability can be linked to SLA guarantees.
The good news is that application performance management capabilities are already integrated into the Cisco IP NGN. These capabilities offer a competitive advantage for service providers over point products such as appliances from vendors that focus only on the WAN edge. Cisco integrated application performance management capabilities can also contribute to greater operational efficiencies and higher profitability as service providers leverage existing network assets to provide new service models and operational views with insight into the end-user experience for each application.
Delivering Application Performance Management
Networks have evolved significantly over the years, but service providers today must still support older, legacy protocols. The coexistence of legacy protocols, platforms, standards, and policies leads to mounting complexity and workarounds that hinder scalability, security, and performance.
Cisco is working to leverage products and technologies in the Cisco IP NGN to reach a new phase, where networks can natively understand the content and context of application traffic and can perform operations on this traffic determined by business policies and rules as it traverses the network. These operations include enhanced security, visibility, marking prioritization, and optimization services for true application performance management within the network.
Application performance management will be pervasive throughout networks, data centers, and service-oriented clouds. It will be a vital capability to hasten the move beyond current point solutions that are limited in scope and applicability, and to evolve towards virtualization and the unbounded mobility of computing resources and applications.
Application performance management within the network will enable:
• Correlation of multiple sources of network instrumentation, such as flow data and Simple Network Management Protocol (SNMP) messages.
• Deep packet inspection to distinguish between and classify different types of applications.
• Application and performance-aware intelligent routing and application prioritization at every point in the network.
• Better awareness of application messages to apply highly granular rules according to business policy.
• ...and much more.
Figure 5 shows the functional components of the Cisco Application Performance Management Framework, with Operations Support Systems (OSS) providing key services across the network.
Figure 5. Functional Components of the Cisco Application Performance Management Framework - APM Functional Components
The Value of Integrated Solutions for the Cisco IP NGN
An integrated approach to application performance management provides many benefits to the managed service provider for true end-to-end functionality, including:
• Simplified deployment of QoS.
• Greater visibility into compressed or noncompressed applications for assigning class of service.
• Enhanced ability to offer incremental services for upselling by leveraging embedded components.
• Easier integration with existing service provider operations by using open-standards instrumentation and carrier class partners for management.
• Tight integration of WAN optimization with visibility and control technologies to help ensure complete feature transparency.
• Reduced total cost of ownership due to leveraging existing infrastructure.
• Ability to correlate static and adaptive network, security, and business policies (for example, support of call admission control for rich media such as streaming video, and the ability to reserve appropriate network bandwidth and security policies on a more granular basis per customer location or service subscription).
• Ability to provide new services that rely on intelligent transport (for example, VPN connection to SaaS and hosted applications).
• Open-standards instrumentation that helps ensure integration with a wide range of existing visibility solutions in service provider networks.
• Consistent implementations of application performance management technologies across routers, switches, appliances, and other gear.
Maintaining an application performance management network requires integration of intelligence in the network. With the application performance management capabilities integrated into the network, service providers can position services that are built on a network environment that is highly aware of applications and their performance, and allows policy-based decisions to be made in alignment with business processes and objectives.
Figure 6 shows tools and technologies available from Cisco that comprise the evolving Cisco Application Performance Management Framework.
Figure 6. Cisco Application Performance Management Tools and Technologies - Managed Service Provider Portal for Application Performance Managed Services
Cisco Network-Based Application Recognition (NBAR) is a classification engine in Cisco IOS® Software that can recognize a wide variety of applications, including web-based applications and client/server applications that dynamically assign TCP or User Datagram Protocol (UDP) port numbers. After the application is recognized, the network can invoke specific services, such as QoS or intelligent path selection, for that particular application. NBAR currently works with QoS features to help ensure that network bandwidth is used efficiently based on defined business policies.
Cisco NetFlow, also available in Cisco IOS Software, efficiently provides key instrumentation to enable services such as traffic accounting, usage-based network billing, network planning, security, Denial of Service (DoS) monitoring capabilities, and network monitoring. NetFlow provides valuable information about network users and applications, peak usage times, and traffic routing. Cisco invented NetFlow and is the leader in IP traffic flow technology. The newest features, Flexible NetFlow and the new industry standard format IP Flow Information Export (IPFIX), enable much greater flexibility and control of information that is exported for any specific service.
Cisco Performance Routing (PfR) helps enable intelligent path selection at the WAN edge based on performance-sensitive routing metrics such as response time, packet loss, path availability, traffic load distribution, cost minimization, and composite metrics representing voice quality. Classic routing mechanisms can provide load sharing and failure mitigation, but Cisco PfR can make real-time routing adjustments based on criteria other than static routing metrics. These real-time routing adjustments based on performance make Cisco PfR an important component for building highly available paths across the Internet and WAN. Because PfR must be performance-aware, it also provides performance monitoring, which can be used as part of an application visibility solution.
IPSLA is a feature included in Cisco IOS Software on Cisco routers that can give network administrators the ability to analyze IP service levels for IP applications and services. It is widely used in networks to generate time-based network and services performance data.
Web Cache Communications Protocol version 2 (WCCPv2), developed by Cisco, specifies interactions between one or more router or Layer 3 switch or web caches to establish and maintain the transparent redirection of selected types of traffic flowing through a group of routers for optimization of resource usage and faster response times.
Cisco Wide Area Application Services (WAAS) is a comprehensive WAN optimization solution that utilizes these Cisco IOS Software features and protocols to accelerate and optimize applications, reduce branch infrastructure costs, improve IT agility, simplify data protection, and help ensure end-to-end delivery of applications that support business processes.
How Service Providers can Position Application Performance Management Networking
With the ability to track application performance, understand application characteristics, enforce policies, and confidently manage applications across networks, service providers can forge new, closer partnerships with customers. The service provider can help IT departments better understand and resolve application performance issues and improve end user and customer satisfaction. Managed application networking services can be more finely tuned to include a company's business objectives and processes.
For example:
• SLAs can be based on application performance instead of network speeds and feeds.
• With increased visibility into networks end-to-end, the service provider can suggest best practices, introduce proactive alerts to reduce oversubscription and downtime, and maintain stringent security.
• Service providers can assist IT departments in network and personnel asset optimization and performance monitoring while offering higher value and higher margin professional services and becoming a trusted advisor to business customers.
Service providers can also utilize application performance management features to optimize their own operations and provide new services, including:
• Offering new incremental revenue-generating services such as VPN with application visibility and monitoring, VPN with managed WAAS, and SaaS with network transport services that offer optimized access to applications.
• Providing the ability, with greater visibility into network devices, to confidently plan the deployment of next-generation service features that may enhance revenue opportunities.
• Quickly isolating anomalies (such as malware, a software bug, or a configuration error), and quickly understanding where a problem originates in the network to better allocate responsibility to either the service provider or the customer's IT department.
• Correlating performance in the network core and edge and in data centers to aid in troubleshooting problems.
Conclusion
Application performance management is becoming a significant business requirement as well as an opportunity for managed network service providers. It is being driven by the increasing globalization of business and the need for collaborative, mission-critical business applications beyond the LAN to the WAN and the borderless enterprise. Application performance management networks are defined by a suite of capabilities that help ensure applications are successfully delivered end-to-end.
Cisco products, technologies, and the Cisco IP NGN architecture provide the features and intelligence necessary for application performance management. The Cisco Application Performance Management Framework provides not only application performance management but many other features for traffic and resource optimization, application visibility, and overall enhanced network control.
The framework allows managed service providers to leverage existing infrastructure while introducing new capabilities that allow for the monitoring and management of applications as they traverse networks and reach users in central or outlying offices or via mobile devices. This departure from the complex, point-to-point application implementations and management solutions that evolved as applications moved from the LAN to the WAN brings new efficiencies and opportunity for managed service providers. Leveraging existing Cisco and Cisco partner products and technologies, service providers can do a better job of providing efficient, secure, and flexible application services to customers. SLAs may now include far more than simply guarantees on speeds and feeds to also encompass application performance that takes into account business objectives and policies.
The Cisco Application Performance Management Framework is evolving quickly to meet customer requirements. Many features are now available that provide the basis for new managed services and operational efficiencies for both service providers and end customers.
For More Information
Contact your Cisco account manager for more information on the Cisco Application Performance Management Framework and vision or visit:
