XLDB - Extremely Large Databases

Conference Program


Printer-friendly version [PDF]
Wednesday, October 6, 2010
08:00 AM Continental Breakfast
9:00 AM Conference Introduction and Logistics
Main objectives, logistics and agenda.
Jacek Becla (SLAC, XLDB4 chair)
9:15 AM Welcome -- Science and Computing at SLAC
Official "welcome" and introduction to new exciting science and computing challenges at SLAC.
Donald Lemma (SLAC CIO and Computing Division Director)
9:30 AM Complex Scientific Analytics at Extreme Scale
A comprehensive overview of how big science approaches complex analytics at extreme scale.
Gregory Dubois-Felsmann (HEP), Andrew Connolly (astronomy), John Caron (atmospheric research), Bill Howe (ocean sciences), Eugene Kolker (bio), Jacek Becla (summary)
10:50 AM Coffee Break
11:10 AM Complex Industrial and Government Analytics at Extreme Scale
A comprehensive overview of how data-intensive industries approach complex analytics at extreme scale, it will highlight similarities and differences comparing to approaches taken by big science.
Irina Vayndiner (MITRE), Steve Hirsch (NYSE Euronext), Mike McIntire (Yahoo!), Peter Breunig (Chevron), Damian Reeves (Quantcast)

moderator: Kian-Tat Lim (SLAC)
12:30 PM Lunch
1:30 PM Operational Issues with Managing Large Database Clusters
Practical, operational issues with managing large database clusters, based on experiences from at least two large-scale industrial setups.
Oliver Ratzesberger (eBay),
Jeffrey Rothschild (Facebook)
2:10 PM Behind the Scenes of Big Science Projects
A talk explaining how big and long-term scientific projects get started, what the decision processes are, how requirements and data volumes are decided, how vendors are evaluated.
Amber Boehnlein (DOE)
2:55 PM Existing Scientific Tools/Formats - netCDF, HDF5, fits, xtc
A set of short talks to make non-scientific communities aware of most commonly used scientific formats and related tools custom-built by scientists.
Daniel L. Wang (SLAC)
3:15 PM Ice Cream Social - Poster Session for Gold Sponsors
3:55 PM Existing Scientific Tools
A continuation of short talks to make non-scientific communities aware of most commonly used scientific tools custom-built by scientists. Considering covering: root, castor, xrootd.
Richard Dubois (SLAC)
4:15 PM Lightning Talks (8 x 5 min)
  1. A (Hypothetical) Data to Discovery Engine
  2. Square Kilometre Array
  3. Exabyte plus initiative
  4. Damasc
  5. Horizontal virtualization on commodity hardware without requirements for database optimization techniques
  6. Array Versioning System
  7. Smorgasbord of Real World Extreme Scale Database Analytics
  8. Introduction to MongoDB
  1. Mark Stalzer, Caltech
  2. Kevin Vinsen, ICRAR
  3. Leon Guzenda, Objectivity
  4. Neoklis Polyzotis, UCSC
  5. Stefan Groschupf, Datameer
  6. Philippe Cudre-Mauroux, MIT
  7. Andrew Lamb, Vertica
  8. William Shulman, MongoDB
5:00 PM Adjourn
6:30 PM Reception and dinner
Thursday, October 7, 2010
08:00 AM Continental Breakfast
8:40 AM Announcements and Logistics
Jacek Becla
8:45 AM Emerging Technologies for Complex Extreme Scale Analytics
Unifying emerging technologies such as map/reduce, streaming databases, and workflow management into a coherent tool set for extreme scale data analytics.
Jeff Hammerbacher (Cloudera)
9:25 AM Emerging Scientific Tools - SciDB
Lessons learned from studying scientific use cases and interacting with scientific communities by the SciDB team. SciDB response to these needs.
Mike Stonebraker (MIT)
9:45 AM Science Benchmark
Introducing the idea behind science benchmark, current status and plans.
Mike Stonebraker (MIT)
science benchmark paper
10:00 AM Coffee Break
10:40 AM Extreme Scale Architectures and New Hardware Trends
Impact of new hardware trends such as solid state disks, GPUs, servers with very-many-cores.
Alex Szalay (JHU)
Tamas Budavari (JHU)
11:25 AM Data Preservation and Integration
Challenges related integrating data from multiple sources and preserving petabytes of data.
Jane Mandelbaum (Library of Congress)
12:05 PM Automated Information Extraction, Content Curation and Machine Learning
Needs in the area of automating data processing and analytics. Perspectives from industry and science.
Raghu Ramakrishnan (Yahoo!),
Kirk Borne (GMU)
12:45 PM Closeout
Next conference planning, final conclusions and closeout.
Jacek Becla
1:00 PM Lunch
2:00 PM Adjourn
Privacy Statement -