Customizing Scientific Data Analytic Platforms via SaaS Approach

Abstract

At the era of data driven science discovery, it is essential to provide customizable scientific data analytic platforms for researchers to conduct their personalized data intensive analysis. Science Gateway has been a viable solution to enabling scientists to run scientific simulations, data analysis, and visualization through their web browsers. But most science gateway frameworks are designed for integrating commonly used software tools and datasets in a specific science domain, thus requiring significant effort to implement the essential variability in lab-specific data processing workflows. In this paper we introduce a multitenancy architecture (MTA) based customization framework that can greatly accelerate the customization cycle of science gateway systems. Each tenant has his own workspace that assembles the software stack and tools to meet the software requirements of his specific data analytics tasks. Through this framework, developers can import their domain-specific analysis pipeline scripts and mashup relevant templates including GUI templates, tool recipes and workspace templates to generate both workspace and web interface for running these application workflows and visualizing the output from workflow executions without writing extra wrapping code.