Information Technology Reference
In-Depth Information
Fig. 3.5:
Example of a Partner Grid
The underlying hardware and operating platforms of a Partner/Community Grid
can be heterogeneous. While the Cluster Grid of one VO member may be Linux/
Intel-based, another VO member's Cluster Grid may be built from a combination of
large symmetric memory Sun servers and storage/backup. This heterogeneity can
be accommodated by a global Grid middleware. The infrastructure of a company
shared in a VO is virtually melted together with the infrastructure shared by its part-
ners. The own infrastructure is opened for common use and therefore security and
trust are important considerations. A VO also requires a certain central control, for
example for monitoring and allocation of tasks. The management of a VO is either
done by a VO member, e.g. the one providing the most resources or having the busi-
ness lead, or by a third party.
Examples
Currently, Partner or Community Grids are mainly used in scientific research. For
example, the White Rose Grid (http://www.wrgrid.org.uk/), based in Yorkshire
(UK), is a VO comprising the Universities of Leeds, York and Sheffield (WRG
2009). There are four significant compute resources (Cluster Grids) each named
after a white rose. The goal of the White Rose Grid is to support collaborative
efforts within eScience research at the involved universities. Two Cluster Grids are
