Software architecture principles for web-application infrastructure – how Earth Science software systems such as the Live Access Server fit in
Roland H. Schweitzer, Weathertop Consulting, LLC, College Station, TX; and S. Hankin, K. M. O'Brien, J. Callahan, A. Manke, S. Du, and J. Li
The Live Access Server (LAS) is a well-established Web-application software system for display and analysis of geo-science data sets. The software, which can be downloaded and installed by anyone, gives data providers an easy way to establish services for their on-line data holdings so their users can make plots, create and download sub-sets in a variety of formats, and compare and analyze data.
One of the most fundamental decisions that software architects must make is how much of their software infrastructure they must write themselves and how much can they leverage from existing open source software frameworks. Often, software engineers working in areas such as the earth sciences feel as though their requirements are too specialized to take advantage of software frameworks designed primarily for business-to-business and on-line commerce transactions and user interactions.
In this presentation we discuss our experiences using off-the-self Web-application frameworks to build the next generation of the Live Access Server. The proper design software components to isolate the specialized processing needed for working with geo-science data from the application frameworks will be a special emphasis of this presentation. In conjunction with the discussion of software infrastructure issues related to Live Access Server development, we will provide an overview of the latest advances in the capabilities and configurability of the Live Access Server itself.
Specifically the new architecture makes it easier to add new output products to our core system based on the Ferret analysis and visualization package. By carefully factoring the tasks needed to create a product we will be able to create new products simply by adding a description of the product into the configuration and by writing the Ferret script needed to create the product. No code needs to be added to LAS to bring the new product on-line. Equally important, these same design principles make it easier to add other analysis and visualization packages (such as CDAT and GrADS) to LAS. The new architecture is faster at extracting and processing configuration information needed to address each request. Finally, the new architecture makes easier to pass specialized configuration information through the system to deal with unanticipated special data structures or processing requirements.
Extended Abstract (60K)
Supplementary URL: http://ferret.pmel.noaa.gov/armstrong
Session 9A, Internet Applications and Cyberinfrastructure
Thursday, 18 January 2007, 1:15 PM-3:00 PM, 216AB
Previous paper Next paper
Browse or search entire meeting
AMS Home Page