The Grid encompasses two areas of
distributed system activity. One is operational with an objective of administrating
and managing an interoperable collection of distributed compute resource
clusters on which to execute client jobs, typically scientific/ HPC
applications.
The procedures and protocols required to
support clients from complex services built on distributed components that
handle job submission, security, machine provisioning, and data staging. The
Cloud has similar operational requirements for supporting complex services to
provide clients with services on different levels of support such application,
platform and infrastructure.
The Grid also represents as a coherent
entity a collection of compute resources that may be under different
administrative domains, such as universities, but inter-operate transparently
to form virtual organizations.
Although interoperability is not a near
term priority, there is a precedent for commercial Clouds to move in this
direction similarly to how utilities such as power or communication contract
with their competitors to provide overflow capacity.
The second aspect of distributed computing
in the Grid is that job themselves are distributed, typically running on
tightly coupled nodes within a cluster and leveraging middle ware services such
as MPICH. Jobs running in the Grid are not typically interactive, and some may
be part of more complex services such as e-science work flows.
Workloads in Clouds usually consist of more
loosely coupled distributed jobs such as map/reduce, and HPC jobs written to
minimize internode communication and leverage concurrency provided by large
multi-core nodes.
No comments:
Post a Comment