The OnlineHPC was a free public web service that supplied tools to deal with high performance computers and online workflow editor. OnlineHPC allowed users to design and execute workflows using the online workflow designer and to work with high performance computers – clusters and clouds. Access to high performance resources was available as directly from the service user interface, as from workflow components.
The workflow engine of the OnlineHPC service was Taverna as traditionally used for scientific workflow execution in such domains, as bioinformatics,[1][2]cheminformatics,[3]medicine, astronomy,[4]social science, music, and digital preservation.[5]
The project motivation is that there is a gap between researcher skills and competence level needed to run high performance computing. There are at least three barriers on the way to HPC:
Researcher needs to find HPC provider and go through procedures to get access;
Researcher needs to install and configure numerous low-level software applications and deal with digital certificates to proceed;
The last requirement stops majority of even the stoutest researchers that passed first two levels. The service aims to reduce the barriers by providing a complete pre-configured set of tools required for work with computer clusters: in-browser terminal emulator, files system browser, credentials manager and massive task tool.
After a while, it became obvious that engineering and scientific tasks require a more elaborate tool suit that enables researchers to execute the flow of tasks – workflows. Unless there is a number of scientific workflow implementations, they are almost all desktop applications and thus the aim was creating an online environment with a focus on simplicity and enhanced user experience.
The project was in use in bioinformatics,[6][7] human health,[8] telecommunications and other domains.
The service also had basic tools to prepare and run massive computational tasks on clusters - in-browser terminal emulator, file system browser and cluster status viewer.
Users could choose to operate existing hardware cluster or to start a configure a new cluster instance in the cloud.