|
Grid Computing and Virtualization
One of the challenges in Grid Computing architectures is guaranteeing that distributed tasks running on the different nodes in the Grid farm
always deliver results which are compatible with the equivalent, original "standalone" process results. Usually, this is achieved by deploying
dedicated virtual servers which load and unload "Images" of the required environment by demand. The downsides of this approach is the
cost and complexity involved in maintaining these servers and keeping the grid image database up-to-date.
XGE takes a different approach to this challenge, guaranteeing reliable
results without the burden of creating and maintaining system images or deploying virtualization software. Uniquely, XGE can also utilize
the idle CPU time of active user workstations connected to the local network, making dedicated servers an option and not a necessity.
XGE Process-Level Virtualization
Tasks running on the XGE Grid are executed as a special "Virtualized Process". The Virtualized Process emulates
the initiating node's environment, file system and registry, causing tasks executing within the XGE Virtual
Environment to always access the initiating node's environment and file system, while utilizing
the remote node's idle CPU and memory resources. In this manner, tasks running on the grid always use an up-to-date version of the
application environment and data.
Tasks executing on a remote node cannot alter the remote node's file system or
registry. Instead, any modification to the file system is carried on on the initiating node,
ensuring the integrity of the distributed process while keeping the remote node's environment
unaffected and protected.
The concept of Process-Level Virtualization means that no pre-configuration of remote nodes is
required in order for an Agent to perform distributed tasks. Instead, all synchronization is
performed automatically and safely by XGE.
|
