Advanced science.  Applied technology.

Search

Investigation into High Volume Data Management and Performance Using Lightweight Client Interfaces, 10-R8886

Principal Investigator
Inclusive Dates 
10/22/18 to 02/22/19

Background

The market for traffic management software has been experiencing an increasing demand for a modern, cross-platform capable, browser-based user interface, commonly known as a thin client. SwRI has previously remained competitive by meeting this requirement using a XAML Browser Application (XBAP), based on proprietary Microsoft® technology. With XBAP technology nearing end of life, this is no longer an acceptable solution. SwRI was unaware of the ability of modern web technologies to meet performance requirements. This project investigated the performance capabilities of a test client using modern web technologies to determine the viability of supporting SwRI’s traffic management software (ActiveITS). Determining options to allow SwRI to effectively address this demand allowed SwRI to remain competitive in the traffic management software arena.

Approach

The project evaluated various architectural approaches and the performance of several libraries, including Open Layers® and Google Maps, when caching data and receiving live updates. A test client was used to benchmark each library under varying amounts of data, update volumes, and bandwidth restrictions. A custom performance monitor tool was used to allow a uniform method of gathering metrics across various platforms and browsers. The platforms tested included Desktop, Pixel 2, and Galaxy S7 devices. The browsers tested included Chrome and Firefox. The performance monitor was designed to gather metrics on memory usage, data retrieval latency, and rendering speeds. These numbers were used to influence and determine architectural approaches that would allow for improved data management and user experience. All testing was initially conducted on a baseline set, equivalent to those used by current ActiveITS customers. Test loads were then increased up to 10 times the baseline set size to identify any limitations across platforms and browsers.

Accomplishments

Key architectural components and performance limitations were successfully identified in the internal research project. Baseline testing established that under all but extreme network conditions, a thin client is capable of handling current ActiveITS data payloads. Furthermore, additional testing identified upper limits on what modern browsers can support, confirming that there is room for growth while still providing a quality user experience. In cases where performance was degraded significantly, due to either platform or data factors, SwRI can now respond to requests for proposals intelligently about these limitations, with less risk, and provide viable alternatives for clients. Important lessons were learned that can inform future development of the ActiveITS Software suite to ensure we allow for an architecture that provides the most opportunity to grow and adapt to changing market and client demands.