You are here
The OVPR’s Business blog recently put up a post comparing the “olden days” of University computing with today. The piece includes insights from MSI Director Jorge Vinals about the differences between computers of the 1960s and those of today.
To learn about MSI’s early days in the 1980s, see the article that appeared in the Summer 2006 issue of the MSI Research Bulletin.
UPDATE: The new storage system will be activated on January 9, 2013. All data should be migrated to the new system during January - July 2013. Complete information can be found on the Panasas Migration Guide webpage.
The Minnesota Supercomputing Institute (MSI) is putting into a production a new Panasas ActiveStor 14 data-storage system.
The new system will consist of twenty shelves of Panasas ActiveStor 14 storage provided by Advanced Clustering Technologies, Inc. and Panasas, and will integrate filesystem and hardware for a single storage solution. The usable storage capacity will be 1.281 PB, and each shelf will be capable of 1.5 GB/sec or 13,500 IOPS. The full solution will be capable of 30 GB/sec read/write and 270,000 IOPS.
The ActiveStor 14 solution will support up to 12,000 simultaneous clients as well as CIFS, NFS, and Panasas's high-performance PanFS filesystem. The solution also incorporates 24 TB of solid-state drives (SSDs) as part of the hardware solution.
The new system will allow users to do things they couldn't do before. As a central storage solution with excellent performance specifications, it eliminate the need to transfer data within systems at MSI while meeting the challenging requirements of users with large sets of data. Current MSI users are required to migrate their data between high-performance scratch storage and capacity project space storage, which has created, for many, bottlenecks in their data flow. The new system will blur the line between these two storage types. Users will be able, for example, to generate data on MSI's flagship Itasca system and use an MSI laboratory queue to visualize it.
Overall, the new storage system will more than double the capacity of MSI's current storage offering. It will enable users both to perform more interesting research with datasets of finer resolution and to operate on enormous datasets that in the past presented operational challenges. Related to these issues, MSI will address the often overlooked need for access bandwidth by employing a capable system of servers for moving data to and from storage, and it will employ a set of data movers to efficiently channel data between MSI users and external consumers and producers of data. Generally, the new storage system will strike a balance between bandwidth and IOPS. Given the explosion of research with large data requirements, the above capabiliites are key to enabling research of the future.
The Panasas ActiveStor 14 solution will give MSI researchers a significant edge in research and establish MSI as a leader in research inolving big data. In addition to being one of the largest storge systems among U.S. universities (1.281 PB of usable storage, 24 TB of SSD, and 1.84 TB of cache), MSI's system will also be one of the fastest among U.S. universities with respect to storage bandwidth.
The system is expected to be delivered in late November 2012.
UPDATE 12/10/12: The system has been delivered.
Further information will be posted on our website and disseminated to MSI users as it becomes available.
Two MSI Principal Investigators are in a group that recently was awarded an NSF Water Sustainability and Climate grant. The grant is for $4.3 million over five years. Professor Efi Foufoula-Georgiou (Civil Engineering, MSI Fellow) is the lead researcher on the grant. One of the other researchers is Assistant Professor Karen Gran (Geological Sciences), from the Duluth campus. The University of Minnesota Twin Cities is the lead institution on this grant.
The research this grant will fund concerns the impacts of land use and climate change on water quality and ecosystem health. It will use the Mississippi River as a prototype and will particularly study the interaction between land use and river network processes.
Professor Foufoula-Georgiou’s research group uses MSI resources to support their work studying precipitation and water systems. Using data received from a number of sources, the group is analyzing and modeling rainfall across large portions of the earth’s surface. The datasets for both the precipitation measurements and for geological features (river networks, channels, etc.) are very large, necessitating the use of powerful computing capabilities and parallel computational capacity. Professor Foufoula-Georgiou is on the faculty at the St. Anthony Falls Laboratory and is the director of the National Center for Earth-Surface Dynamics.
Professor Gran uses MSI in her studies of the sediment in the Le Sueur River Basin in support of work by the Minnesota Pollution Control Agency, which seeks to evaluate and control the condition of the river. The Gran group studies the sediment load of the river by taking measurements in various ways. MSI is used to store large spatial datasets used in the development and execution of the sediment routing model and to facilitate data sharing between researchers. This work was included in the MSI Annual Research Highlights 2011.
More information about the NSF grant can be found in the UMNews news release.
MSI Fellow Efi Foufoula-Georgiou (Civil Engineering) has been named by President Barack Obama to serve on the U.S. Nuclear Waste Technical Review Board. The Board is an independent agency whose purpose is to provide independent scientific and technical oversight to the DOE’s program for managing and disposing of radioactive waste and spent nuclear fuel. More information can be found in the University’s news release.
Professor Foufoula-Georgiou has been an MSI researcher for many years, and she has been an MSI Fellow since 1996. Her research group uses MSI resources to support their work studying precipitation and water systems. Using data received from a number of sources, the group is analyzing and modeling rainfall across large portions of the earth’s surface. The datasets for both the precipitation measurements and for geological features (river networks, channels, etc.) are very large, necessitating the use of powerful computing capabilities and parallel computational capacity. Professor Foufoula-Georgiou is on the faculty at the St. Anthony Falls Laboratory and is the director of the National Center for Earth-Surface Dynamics.