By Rob Mitchum // September 1, 2016
As more scientists turn to computational methods to work with larger and more complex sets of data, there’s a greater need for computation to meet these researchers halfway. Not everyone has access to a supercomputer or cluster, and even if they did, they might not have the skills to easily transfer their work to those resources.
In late 2014, the National Science Foundation (NSF) took a major step to lower this barrier of entry by funding Jetstream, a new cloud infrastructure that dramatically shortens the on-ramp for computational research. After 20 months of development and testing, Jetstream is now ready for its debut, officially launching full operations on September 1st.
The Computation Institute joined several academic and industry partners in bringing Jetstream to fruition, including frequent collaborators such as the Texas Advanced Computing Center (TACC) and the Indiana University Pervasive Technology Institute (PTI). The CI’s Globus provides the user authentication or login mechanism for Jetstream, the first step in linking users to a broader universe of powerful resources such as XSEDE, the Extreme Science and Engineering Discovery Environment.
“Jetstream is one of the first large-scale NSF systems to use this particular mechanism to integrate with XSEDE,” said R. Lee Liming, a technology analyst with the CI and Argonne National Laboratory, and the University of Chicago representative on the Jetstream technical team. “It also means we have a new group of researchers who are using Globus on an everyday basis.”
The primary goal of Jetstream is usability, making high-performance computing accessible from laptops, tablets, and desktop computers. Instead of using advanced tools such as the command line, SSH clients, and UNIX to interface with a remote resource, Jetstream lets users login to a web interface and create their own virtual machine with the software, data, and computational power they need. Users can also save these virtual machines for later, share them with colleagues who do similar work, or request a VM tailored to the use of a particular community, such as genomics or psychology.
“Jetstream demonstrates the use of cloud computing to expand access to high performance computing resources, tools, and services, which are increasingly vital in advancing scientific discovery and innovation,” said Irene Qualters, head of Advanced Cyberinfrastructure at NSF, in an Indiana University news release. “Today’s researchers and educators require easy access to a rich portfolio of computing and data-intensive resources, tools and expertise to advance science and engineering frontiers. Jetstream users will benefit from cloud capabilities which are integrated into the national research infrastructure.”
Jetstream is one of the first large-scale uses of Globus Auth, a set of services that enable easier authentication and authorization of users. Originally designed to allow researchers to use their campus credentials to access Globus data management services, it will now enable the same experience with Jetstream and XSEDE. The same credentials can be used to access an even broader range of resources, including some that will launch after Jetstream’s official September 1st debut.
“We believe that cloud computing is going to be very important for the future of science, and it is going to open up the NSF resource portfolio to a much wider audience across the country,” Liming said. “We’re expecting a whole new type of scientist will be able to take advantage of our national resources now.”
Additional partners on the Jetstream project include University of Arizona;; Johns Hopkins University, the University of Texas at San Antonio; and Cornell University. Collaborating partners are Jackson State University, the National Snow and Ice Data Center, the Odum Institute at the University of North Carolina, and the University of Hawai’i. Vendor partners include Dell Inc. and the Mathworks.