There are some undeniable truths about the state of IT operations today and their direction for the future. The recent recessionary environment, coupled with IT talent shortages that already existed, has turned the IT operations group into an embattled organization - greater and greater responsibilities with fewer and less-trained resources to manage very complex, n-tier business systems. IT administrators need all the help they can get from the Enterprise Systems Management vendor community; not just more monitoring software ("Yes I can see it, but what does it all mean?"), but rather an intelligent and active management approach.

Applications that are crucial to business success have been moved online. This online trend has been driven by business objectives like generating and protecting revenue or improving margins by moving customers to more efficient Internet-based channels. Especially for mission-critical online business systems, there is the confounding circumstance of highly complex, multitiered IT environments that, as a result of explosive surges in transaction volumes and continual technological advances, have become monumentally more difficult to manage. A typical system could include firewalls, routers, HTTP load balancers, and gigabit ethernet switches at the network layer, with Web server clusters, multiple application server clusters, message-oriented middleware (MOM) servers, legacy gateways, and mid-tier database clusters for the application infrastructure. To make matters even more complex, the added complication of component programming is that the way you deploy components (deployment descriptors) affects performance and availability of your application as much as the quality of the code itself.

Improving application integration has also become an essential component of contemporary IT strategy. In a recent Morgan Stanley survey, 225 CIOs ­ a clear 80% ­ indicated that they would begin new application projects in 2002 with application integration as their top priority (eclipsing other hot topics such as security and network capacity). Application integration has become a top priority as companies experience escalating economic pressure to mine additional productivity from existing, disparate but related applications, legacy systems, and new Web services, while ensuring that they all function faultlessly together.

Where there is challenge, there is opportunity. The challenge is substantial ­ complexity is making it more difficult to manage any business system as a cohesive computing entity, creating the possibility for revenue loss through dropped transactions and expense overruns as customers revert to more expensive offline channels and IT personnel take exponentially more time to fix infrastructure problems. Unfortunately, status quo for most companies, by either design or legacy, is to have separate monitoring tools for each component. This requires that a human being review the output across the different servers and sub-components to determine what is merely a symptom and what is the actual cause of the problem. Customers and analysts agree: many of the current management solutions remain part of the problem, since they actually increase time-to-resolution by deluging users with hundreds or thousands of datapoints ­ rather than answers.

The opportunity: find a tool that reduces the time and difficulty experienced when managing a complex n-tier environment. Analyst firm Enterprise Management Associates believes that real-time management solutions need to meet three requirements to be effective. First, the solution must apply across all the interrelated components in real-time via some form of automated correlation. Second, it must bring immediate value with embedded expertise on how to fix the performance or availability problem. Third, it must account for the reality that all implementations include customized applications and IT configurations.

Clearly, that would require a management solution that can automatically detect problem situations across the multiple tiers and subcomponents (domains) of the business system infrastructure, quickly isolating the fault and suggesting exact fixes.

Even better, what if the management software could actually invoke those fixes or preventative actions across the different components once problems are detected? An intelligent and active management approach like this would ensure that online revenue continues flowing and would considerably reduce the difficulty in managing these infrastructures and allow companies to leverage their stretched IT resources more effectively. This is truly the blueprint for the "next-generation" applications management solution so desperately needed right now.

The true next-generation applications management solution provides immediate cross-domain fault and performance isolation and resolution. Problems as varied as excessive network latency between clustered application servers or purchase orders held up in an asynchronous message queue can be automatically fixed or prevented when detected, resulting in greater online transaction throughput, direct expense reductions, and drastically reduced downtime.

Implementing a next-generation J2EE management solution would produce immediate, significant, measurable improvements. It¹s the only real way to maximize your J2EE performance and substantially accelerate the ROI on your investment. It would be like finding a pot of gold for your company.

• Telco Groups Sitting on Pots of Gold
• Semantic Interoperability: The Pot of Gold Under the Rainbow