The City of Boise is a local government in the capitol city of Boise, Idaho, USA. The City employs approximately 1,500 permanent employees.
The Human Resources Department is preparing to launch an employee self-service benefits application. During the open enrollment period, we expect the number of concurrent logins for this application to exceed 200.
We needed to establish a performance baseline for our current system configuration. Once this baseline was understood, the business could intelligently determine if expansion of the system’s configuration is warranted.
The IT members of the project team did not have any way to determine how the system would perform at very high levels of activity.
The application development staff had no tools to accomplish automated load testing. Earlier projects employed testing teams from the business to create runtime workloads.
Application tested: The application that was tested with NeoLoad is an Oracle Peoplesoft eBenefits self-service system. Each year, employees are required to choose their benefit options within a 3 week period. This new self-service application moves the process from paper to online forms. Acceptable system performance is crucial to the success of the product. The IT staff had no real idea how many concurrent logins the system configuration could support.
- Static HTML web pages served by our Microsoft IIS 6.0 server
- Dynamic HTML web pages served by BEA WebLogic 8.1
- BEA Tuxedo 8.1 throughput and CPU loading
I chose NeoLoad because of its intuitive design and feature rich functionality. Our IT staff does not have much experience with load testing tools. Evaluating other Java-based tools on the market, it was clear that NeoLoad would enable us to conduct effective load testing in the short amount time we had available before the system launch.
The IT staff determined that we needed to establish a performance baseline within the 2 week window of time and still allow adequate time to expand the system’s configuration before the product launch if required. Our objectives were to determine how many employees could be concurrently electing benefit options online while the system maintains a <3 second response time per page.
Test Results and Added Value
The major benefits from using NeoLoad:
- A performance baseline was established for the system configuration
- Confidence levels of the IT staff were greatly improved knowing what the performance level was
The results of load testing:
- We determined that our current system configuration can support 250 concurrent sessions while maintaining <3 second response time experience for the employees;
- We identified equipment in our network that created a performance “bottleneck” during high levels of system activity;
- NeoLoad identified web pages that were inefficient;
- NeoLoad identified TCP/IP configuration issues on both our WebLogic and Tuxedo servers.
NeoLoad provided excellent performance reporting which was used to give feedback to the business.
NeoLoad was an indispensable tool when it came to testing our high risk self-service application. Because of NeoLoad, we found unknown issues and had time to resolve them before the product launch date. I am confident that our benefits application is a better product today and will completely satisfy the expectations of our customers.
Given its ease of use, quality design, thorough reporting, and flexible pricing model, NeoLoad will be the go-to-tool to identify performance and design issues in Citywide and public applications that we design.
Additional comments: “I very much appreciate the staff at Neotys. When we found a device in our network that needed to be upgraded, Neotys graciously gave us a license extension that allowed us to retest the system once the new equipment was installed.”