Pattern mapping in scientific workflows to smart pipes in filesystems
Abstract
With the rapid rise in computing technology the complexity of scientific processes has also increased. Today scientific computing has become very data intensive and relies on scientific workflow systems. The scientific processes don’t have any kernel level support for workflows. A file system which can provide kernel level support to address the problem of parallel processing and portability can improve the overall performance of scientific processes. This thesis focuses on integrating a scientific workflow management system with such a new file system to improve the performance of workflow management systems for scientific computing.
Degree
M.S.
Advisors
Hacker, Purdue University.
Subject Area
Computer science
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.