In-Depth

Optimizing Application Delivery in the Extended Enterprise

An application delivery platform suite can automate its end-to-end deployment

• By Nat Kausik
• 11/30/2004

As organizations increasingly rely on Web applications to support distributed offices, partners, customers, and “road warriors,” IT departments are pressured to fulfill the enterprise’s growing need for continuous information flow across the extended enterprise while staying within an already strapped IT budget.

It has become all too common for organizations to sink millions of dollars into Web-enabled applications only to realize that post-deployment performance issues seriously inhibit user adoption. When Web applications aren’t performing, productivity goes down, IT expenditures increase, and dissatisfied customers, partners, and employees are reluctant to adopt them.

In most instances, the problem is rooted in the gap between the conceptualization and development of Web applications, and the realities of the deployment environment. Bridging the gap is the key to optimizing application delivery. When development teams are separated from the specifics of the deployment environment and the eventual operation and delivery of the application, post-deployment issues manifest in slow user response times, poor application adoption and other factors that contribute to the business bottom line.

To compensate for this and address the unique challenges involved in ensuring maximum application service delivery to any user, anywhere, IT proactively seeks solutions that can ensure the most effective delivery of business-critical applications. A category of technologies has emerged to meet this need: Enterprise Application Delivery Solutions. These technologies take over where development left off, optimizing the application’s functionality for its specific deployment scenario and providing an automated end-to-end environment that includes all the support functions necessary for an organization to deliver high application service levels across the Extended Enterprise.

The Need for End-to-End Automated Application Delivery

In pursuit of application performance gains, many IT departments undertake labor-intensive manual adjustments to applications, or install costly network equipment to increase bandwidth capacity. Further, they spend countless hours managing application upkeep and security through piece-meal patch management, and allocate even more budget dollars to monitoring technologies that provide limited, synthetic estimates of the service levels applications are delivering to end-users.

There are several intricately linked factors involved in managing applications in the extended enterprise, and approaching them in such a fashion leads to wasted resources, IT headaches, and overly-bloated IT systems. To successfully deliver consistent application service to users at any access point, IT must take a holistic approach that addresses the end-to-end process of application delivery, including application performance, end-user experience monitoring and application security. Application delivery platforms provide a single environment that integrates all of these functions, driving down the cost and time involved in managing the Extended Enterprise while boosting application service levels. By leveraging both the system and network infrastructure, application delivery solutions provide an integrated view of the application from the database through to the end-user.

Enterprise Application Delivery Specifics: Optimization, Monitoring, and Security Integration

Traditionally, organizations address the various components of application delivery —performance optimization, monitoring, and security—in silos by purchasing technologies and investing in processes that address each one separately. This significantly increases the complexity of the Web application infrastructure and often fails to yield significant improvements.

When applications that perform well at corporate headquarters slow or sputter in branch offices, many organizations immediately throw bandwidth at the problem, making significant network infrastructure investments. Application delivery solutions understand and address all factors that can affect performance in a WAN environment, including server overload, application inefficiency, latency, and application prioritization—issues that more bandwidth will do nothing to solve.

For most companies, the second solution silo is monitoring. Nearly every company that has incorporated Web applications into its business has made some level of investment in a performance-monitoring tool. Typically, such solutions substitute the user-experience simulations (thanks to emulation technology) for true visibility into what the user is experiencing.

Going this route adds yet another layer of complexity and separates performance optimization from troubleshooting. A more effective model, enabled by application delivery platforms, is to create a tightly integrated closed-loop between monitoring and optimization that enables a company to track performance optimization results over time, identify trends, and reduce the overall IT costs associated with problem identification and resolution.

The final silo in the traditional application delivery scenario is Web application security. For many organizations, this means engaging in a “patch management” race against relentlessly ingenious hackers. The cost and complexity of this never-ending task is steep. By incorporating an application firewall—an additional layer of security that polices the application layer—an application delivery platform can cut down on the “patch scramble” and integrate the ongoing task of protecting applications with the other critical management functions involved in maintaining the Extended Enterprise.

Single Solution, Multiple Benefits

To put this all in context, consider the case of one Fortune 500 financial services corporation that handles risk and wealth management for more than 50 million clients and employs well over 100,000 employees worldwide.

The company recently launched a key business-automation initiative to consolidate business processes around two Web applications: a packaged CRM solution and a custom J2EE application to manage portfolio and risk-management advisor functions.

Upon rollout, branch office users found the applications to be painfully slow, while performance at headquarters was marginally acceptable. To complicate matters, IT had no visibility in to the actual experience of the users at the branch offices. In this very common scenario, most companies take a similar multi-step approach to addressing the performance problems, commencing the process by making investments to beef up the network infrastructure or replicate the data center.

In many cases, step two involves purchasing performance-monitoring tools from a point vendor in order to determine the source of the degradation. Finally, because of the sensitivity of financial data and compliance requirements, the IT department would have already invested in a separate security solution that may be hindering performance. In addition to the labor and time involved in integrating these solutions, IT may very likely be called to the mat to justify the additional investments in what is already a pricey automation initiative. Finally, the complexity inherent in this solution model will add to the total cost of ownership of the IT asset through increased maintenance costs over time.

Ultimately this financial services company rejected the traditional route, opting to implement an application delivery platform suite to automate the end-to-end deployment of these two applications. By purchasing a single solution that could be installed in the data center at headquarters, the company was able to dramatically improve end-user response times for branch offices while decreasing the applications’ bandwidth usage by nearly 90 percent. Further, the solution prevented the company from having to make significant additional investments because it optimized, monitored, and protected both Web applications from one common environment while leveraging the existing infrastructure.

Optimizing the delivery of Web-enabled application can reduce infrastructure costs, reduce performance-engineering costs, enhance IT productivity, and improve user experience. Tying together the various application-support functions and automating the process of optimizing performance, monitoring and managing security is a highly accessible way that any IT organization can begin doing more with less. Taking practiced, proven measures to reduce complexity of the enterprise while boosting service levels ultimately lowers the total cost of ownership of the application.