| |
Physics experiments and simulations grow in size and complexity. Examples are the existing HEP/NP experiments and upcoming challenges of SNS, LHC, ILC, and ITER. Managing the experimental data is an extremely complex activity. Physics simulations now attempt full modeling of various phenomena and whole experimental devices, such as in fusion integrated and space weather modeling. Recent advances in computer science, such as Grids and Components, address the challenges faced by applications. In science, Globus and Common Component Architecture (CCA) became commonly used tools for these technologies. Globus allows creating a grid–computers trusting each other and a group of users who can then submit jobs and move data. CCA expresses connectivity of the simulations elements in different languages as “components,” objects with in and out “ports.” CCA “frameworks” combine components into simulation and can swap components sharing ports. CCA accommodates high-performance and distributed applications. We will present our work with Globus and CCA in HEP/NP and fusion, share the lessons learned, and evaluate the ease of using these technologies and the value added.
|
|
