Information Technology Reference
In-Depth Information
# of images processed by day
800000
700000
600000
500000
400000
300000
200000
100000
0
21/6/07
5/7/07
19/7/07
2/8/07
16/8/07
30/8/07
13/9/07
27/9/07
FIGURE 3.7
DILIGENT data challenge.
content-based data retrieval and automatic data classii cation techniques
that combine both text and image content, expanding the limits of conven-
tional search engines, which can only search text associated with images
and audio-visual content.
Figure 3.7 shows the images processes per day. Three phases can be
identii ed, separated by vertical lines in the i gure: the i rst phase was
used to test the overall setup, particularly interactions with the Flickr
database and the EGEE infrastructure; in the second phase, EGEE resource
centers have been gradually added until all resource centers supporting
this work (10) were able to contribute, which is when the phase third
started. The negative dips in the i gure correspond to times where major
upgrades in the software have been performed.
3.5
The vision of grid computing that launched EGEE implies a federation of
resources across institutional, disciplinary, and national boundaries, in
contrast to “enterprise grids” that often exist within individual companies.
The broader vision pursued by EGEE and others requires members of indi-
vidual grids to be aware of and cooperate with other related grid efforts to
work toward interoperability at both European and global levels. Hence
EGEE is working closely with other grid projects in Europe and worldwide
to implement the vision of a seamless worldwide e-Infrastructure. These
projects include the US-based Open Science Grid (OSG, http://www.
opensciencegrid.org) and TeraGrid (http://www.teragrid.org) projects, the
Japanese NAREGI (http://www.naregi.org) project, and the European
DEISA (http://www.deisa.org) project, a number of projects that extend the
EGEE infrastructure to new regions (see Table 3.3 ) as well as (multi)national
efforts like the UK National Science Grid (NGS, http://www.ngs.ac.uk), the
The Worldwide Grid Landscape
 
Search WWH ::




Custom Search