Based on these unique requirements and the solutions considered, we decided that Application Service Modeling was the best approach to meet our needs. Application Service Modeling provides the necessary contextual visibility. It dynamically correlates the services provided by applications to the underlying code components supporting those services providing the necessary contextual visibility to manage their performance effectively. A recent Forrester Research report said that "the model-based approach is a ‘must-have' for more complex composite applications, such as SOA apps and portals."

Application Service Modeling Enables Superior Service Delivery
Model-based monitoring of service-oriented applications is all about increasing contextual visibility. Application Service Modeling delivers visibility into our service-oriented applications in the context of which business services are being enabled by which application components across our SOA infrastructure. Once IT had the necessary contextual visibility, we were able to manage one or more components to tune service delivery.

Modeling at the application level returns a visual topology model of the infrastructure under each application service, so the dependencies integral to providing each service are laid bare. How this works is that the metadata in jar files, configuration files, and databases are parsed to detect all entry points. After this procedure is completed, the services are modeled, the metrics are determined, and then both are overlaid on the model.

Any models that the tool generates automatically update themselves as underlying components and relationships change, and do not require manual intervention or attention to these updates. Automation gives you a much better chance of keeping your business-critical services up and running. Additionally, IT organizations need a dependable means for doing an impact analysis. For example, before a pre-planned change is made, IT operations must be able to identify all application processes, application components, systems, etc. that will be impacted by the change. This adds significant value in a way that's not possible without having the necessary context to bridge the IT visibility gap.

Application Service Modeling provides a degree of visibility and troubleshooting capabilities we simply did not know was possible. For example, the system revealed in five minutes the root cause of a problem on which we had spent eight weeks trying to diagnose. So what was the problem?

A search function on our Web site that allows customers to find and purchase products wasn't working - a big issue to say the least. My staff spent six weeks trying to nail down the root cause. The job was complicated by all the third-party software that we had either acquired or integrated into our data center. We launched a new version of a client's commerce site, but didn't see the issue arise until the transaction volume rose, and then it impacted us across our whole Web tier. At first we assumed it was a database issue with connection pooling, specific to our application.

Within 24 hours of deploying Application Service Modeling, we were able to trace full transactions not just in the systems we had developed but into third-party tiers and applications as well. That in turn allowed us to isolate the problem. This was significant, since we used to spend so much time debating with vendors who needs to fix a problem. Service modeling is what allowed us to tune and test, over-tune and test while monitoring for the optimal performance fix. Doing SOA right demands an accurate diagnosis of trouble sources, especially across virtualized and shared resources.

Impressed by the capabilities of an Application Service Modeling approach to diagnostics and repair, we have adopted the technology across our business-critical commerce sites domestically and worldwide. The benefits realized have included shorter mean time to repair, improved application service levels, and greater return on investment for our service-oriented applications. We identified additional areas for improving the central Web commerce application code leading to greater stability and improved performance. One of our SOA developers came to believe that employing Application Service Modeling earlier could literally have saved years of development time (I'm not kidding). But perhaps most important, it helps us fulfill our central mission of earning the respect and loyalty of our business partners through superior value and service.

Conclusion
Application Service Modeling bridges the IT visibility gap inherent in today's dynamic composite applications. Without having this contextual visibility, it has become impossible to effectively diagnose application service problems in enterprise applications that leverage multiple layers of middleware under a barrage of constant change. An unmanageable application, no matter how sophisticated or rich the services, provides little value to an organization. Further, an IT operations team must understand an application's logic to be able to effectively conduct the impact analysis required to stay ahead of potential application problems. By managing complex composite applications at the application services level, IT organizations can maximize application performance levels, reduce the mean time to resolution, and maximize productivity. A more sophisticated approach is needed to achieve the "last mile" of aligning IT with the business and enabling the organization to yield expected ROI from its composite application investments.

Resources

• www.IngramMicro.com
• www.ClearApp.com
• www.EnterpriseManagement.com