http:wwwconf.slac.stanford.eduxldb07 - PowerPoint PPT Presentation

1 / 27
About This Presentation
Title:

http:wwwconf.slac.stanford.eduxldb07

Description:

... science & industry (pattern discovery, multi-d aggregation, unpredictable ... 1 year, 2-3 days, _at_SLAC. Don't expand size much. By ... Two Days, Three Goals ... – PowerPoint PPT presentation

Number of Views:49
Avg rating:3.0/5.0
Slides: 28
Provided by: ncr50
Category:

less

Transcript and Presenter's Notes

Title: http:wwwconf.slac.stanford.eduxldb07


1

Welcome!
2
  • http//www-conf.slac.stanford.edu/xldb07

3
(No Transcript)
4
One Day, Three Goals
  • Identify trends and major roadblocks related to
    building extremely large databases
  • Bridge the gap between users trying to build
    extremely large databases and database vendors
  • Understand if and how open source projects like
    the LSST Database can contribute to the previous
    two goals in the next few years

5
Things We Talked About
6
  • Valuable data discarded due to scalability limits
    and cost

7
Substantial commonalities between science
industry (pattern discovery, multi-d aggregation,
unpredictable query load, procedural language
needs, )
8
Industry leading scale, science leading
complexity of analytics
9
Parallel, shared-nothing architectures on
commodity clusters are becoming very popular
10
Roadblocks funding problems, vendor-users
disconnect, science-academia disconnect
11
Rebuilding, not reusing software
12
Gap between needs and what vendors offer is
widening
13
Structured and unstructured data coming together
14
MapReduce popular, but lacks efficient joins
15
Things We Decided
16
?
  • Conduct another workshop in 1 year, 2-3 days,
    _at_SLAC
  • Dont expand size much
  • By-invitation only
  • Focus on experience sharing, commonalities that
    can be developed into community-wide requirements

17
?
  • Try to setup smaller workshop and/or working
    group(s)
  • In particular science db academics

http//xldb.slac.stanford.edu/display/XLDB/SciDB
18
?
  • Set up shared infrastructure
  • Initially wiki, possibly test-bed environments

19
  • Try to define a standard benchmark focused on
    data-intensive queries

20
  • http//www-conf.slac.stanford.edu/xldb08

21
Two Days, Three Goals
  • Continue to understand major roadblocks related
    to extremely large databases with an emphasis on
    complex analytics
  • Continue bridging the gaps within the XLDB
    community including science, industry, database
    researchers and vendors
  • Build the open source SciDB community

22
It Is All About Ad-hoc Discussions
  • You are expected to speak up too
  • But no sale speeches, please
  • Discussions are not electronically recorded
  • Detailed report will be released
  • Once OKed by workshop participants

23
Attendance Rough Breakdown
24
Attendance Rough Breakdown
If this group wont make a difference, who will?
  • Big science
  • Big industries
  • All major DBMS vendors
  • Very promising startups
  • World-class DB researchers
  • Superstar DB programmers

25
Dinner
  • Location
  • Sheraton Palo Alto
  • Driving directions available
  • Reception
  • 700 pm 730 pm
  • Dinner
  • 730 pm 1000 pm
  • Buffet
  • Cost
  • Free
  • Maybe except the valet parking

Make sure you wear your XLDB2 badge
26
BIG Thanks to Our Sponsors
27
Agenda
http//www-conf.slac.stanford.edu/xldb08/agenda.ht
m
Write a Comment
User Comments (0)
About PowerShow.com