The thesis describes a web application that offers users an easy way to run predetermined analyses in a HPC environment without having to use a command-line interface. The goal was to create a web application that simplifies the use of commonly run analyses inside a HPC cluster. We review existing solutions to this problem. We then provide a high level description of an application that allows users to manage input data, metadata, analyses and output data as a graph. We describe a conceptual model for this application and its constituent parts. In the following part we closely examine the inner workings of the implementation. We finish with an example analysis running in a test environment. Finally, we describe the application's limitations, possible upgrades and improvements.
|
