Wily Solution Scenario — Proactive Application Management

A leading financial services institution is rolling out a brand new Personal Asset Management (PAM) application for its corporate customers. The application will be used by the employees of its corporate customers to analyze their personal portfolios and allocate assets using different financial products such as stocks, mutual funds and bonds. Running on a Web Application Server, the new application is expected to transact millions of dollars a day and support a peak load of 10,000 concurrent users. It also communicates with back-end systems including DB2 and various CICS regions.

The new PAM application replaces one that failed to meet the performance standards defined in the firm’s Service Level Agreements (SLAs): 90% of requests were to be completed within one second. The PAM application is expected to meet or exceed this performance level. The firm’s Operations and Application Support Group (OASG) decided to take a team-based approach to managing the PAM application and developing Java application management best practices. They have chosen the Wily solution to proactively manage the whole application and to maintain high availability in the production environment.

The OASG requires that every new application be thoroughly tested in QA prior to deployment. The QA team stress-tests the application and uses Wily’s Introscope® to monitor its performance from the inside, at the component level. They find that the application’s response time spikes significantly after 100 hits per minute, well short of the threshold required to support its peak usage. Introscope reveals that a specific method inside a frequently used EJB is logging too much information, thereby causing excessive I/O latency. Armed with this information, the Application Development team fixes the problem quickly.

Additionally, the team uses Introscope to measure baseline performance, such as maximum users and maximum transaction rate. These measurements are used to establish thresholds for Introscope’s built-in alerting system. The QA team determines that the application meets all pre-production performance criteria and is ready to be deployed.

Once the application is deployed, the OASG uses Introscope around-the-clock (24x7) to monitor the PAM application and all the critical components inside it that execute customer transactions. Introscope’s free-form dashboards allow each member of the OASG to monitor live data that is most meaningful to him or her. The Ops Manager’s dashboard provides him with a bird’s eye view of the application infrastructure and all its supporting systems such as Web Servers, Application Servers, Database Servers etc. The Application Support specialist’s dashboard monitors the components inside the application and filters out the worst performing components in each category (e.g. Servlets). The Application Server administrator’s dashboard monitors the resource utilization inside the Application Servers and alerts the admin of potential resource depletion. Additionally, Line of Business (LOB) managers use Introscope dashboards to monitor total transactions executed on a daily basis and the total number of Web site visits during a particular time period to measure the success of marketing campaigns.

In addition, they have chosen a number of Introscope extensions to broaden their visibility into the entire Java application environment. They use Introscope PowerPacks™ to monitor critical application server resources (e.g. HTTP Sessions, Execute Threads, JDBC Connections Pool) as well as the interaction between the application and the different CICS regions. The OASG uses Introscope’s Environment Performance Agent (EPA™) to monitor the health and availability of supporting systems including Web Servers and Database Servers.

Introscope Alerts will instantly notify the team if specific user transactions exceed predetermined performance thresholds. For example, if a specific Servlet (say Login) exceeds a 5 second response time, Introscope notifies the OASG team. Additionally, OASG uses Introscope to validate the service levels of a third

party transaction server at a clearing house by monitoring the RMI calls, complete with thresholds and Alerts. The CIO and Operation Manager use Introscope’s persistent data to generate reports to track Service Levels and QoS targets.

This sample overview dashboard provides a bird’s-eye view of the application and its supporting systems. Learn more about Introscope...

Introscope PowerPacks offer out-of-the-box visibility into critical application server resources. Learn more about PowerPacks...

Solutions .. Products .. Services .. Support .. News & Events .. About Wily