TechTalks Event
Managing Massive Data Storage: What Are the Ways?
with guest expert Anurag Shankar of Indiana University
May 2, 2002
Audio
• Streaming
MP3
• Download
MP3 (Download
Tips)
Why do we need massive data storage? How big does it have to be before it is considered "massive"? (In other words, how big is big?) What kinds of data need this type of special technology? How is today's massive data storage problem different from about ten years ago? How many ways of doing massive data storage are there? What are the technologies involved? What are the reasons behind these technology choices? Where do SANs fit into all this? Where is it all going? Is there a massive data storage system in your future?
Mark Bruhn and Judith Boettcher quizzed our distinguished guest expert - and asked your questions, emailed in ahead of time and during the Tech Talk, to expert@cren.net.
Guest Expert
Dr.
Anurag
Shankar is Manager of the Distributed
Storage Services Group in the University Information Technology
Services Department - and Adjunct Assistant Professor in the Department
of Astronomy - at Indiana University.
Anurag has supplied some interesting links, below. This one is so much fun, we had to put it up here so that you would be sure not to miss it!
Co-Hosts
Mark Bruhn (at left), also of Indiana University, is standing
in for Howard Strauss (top) of Princeton University this week as
TechTalk's Technology Anchor. Judith Boettcher (bottom) is
CREN's Executive Director.
Together, Mark and Judith will ask the really tough questions—and relay the questions you email to them at expert@cren.net.
Background & Resources
Pictures of Massive Data Storage Hardware at Indiana University
- Animation of the construction of a StorageTek 9310 "Powderhorn" tape silo at Indiana University Purdue University in Indianapolis.
- Picture of the StorageTek 9310 "Powerderhorn" tape library (courtsey StorageTek).
- Pictures of the IBM 3494 Tape Library at Indiana University-Bloomington.
Indiana University uses HSM - Hierarchical Storage Management. Here's a white paper for some background on that topic from Computer Associates.
Mark intends to reference An Architecture for New Modes of Science in his opening remarks. It's from the San Diego Supercomputing Center.
Here's a set of good links from CERN, relating to its HEP Mass Storage (HEPMSS), one of which is the IEEE Storage Systems Standard Working Group. Anurag has shared a couple of recent documents of his with us:
- The first is a Word document from the proceedings of the 10th
NASA Goddard Conference on Mass Storage Systems and Technologies,
College Park, Maryland, earlier this year: Building
a Massive, Distributed Storage Infrastructure at Indiana University.
- The second is his
PowerPoint slide show from the same presentation.
The following are live hyperlinks to the references at the end of his Word document, above.
- Information about the
High Performance Storage System (HPSS)
- IBM's DCE
website
- IBM's Transarc
Labs DCE/DFS website
- Samba project
website
- Netatalk
project website
- Apache
project website
- Paul Henson's mods
for Samba/Netatalk/Apache
- Indiana's high-speed
research network website
- Information about IU's
distributed storage services
- Indiana's distributed storage services group website
- Information about the
High Performance Storage System (HPSS)
The Texas Advanced Computing Center at the University of Texas.
The Slice File System at Duke University - and a list of related publications. Also at Duke, a Pilot to Supporting Computational Aspects of Science and Engineering.
From the Australian National University, the Mass Data Store User Guide.
