System and method for the allocation of grid computing to network workstations

In a computer network including at least two workstations and implementing a grid for executing a grid workload on the shared resources of the at least two workstations, a method of assigning said grid workload to one of the at least two workstations, the method comprising:

• monitoring a utilization level indicative of the extent to which a local workload is utilizing the resources of a first workstation of the at least two workstations and setting a first workstation activity state to;
• idle if said utilization level is below a first threshold, or to
• active if said utilization level is above said first threshold and said first
• workstation activity state is already idle or active, or to
• blocked if said utilization level is above said first threshold and said first workstation activity state is already blocked, and;
• checking said first workstation activity state;
• in response to a first workstation activity state of idle, assigning said grid workload to said first workstation;
• in response to a first workstation activity state of active, determining a first workstation preference for accepting said grid workload when said first workstation activity state is active;
• in response to a determination that said preference is yes, assigning said grid workload to said first workstation;
• in response to a determination that said preference is no;
• setting said first workstation activity state to blocked;
• holding said grid workload for assignment to another of the at least two workstations;
• and
• in response to a first workstation activity state of blocked, holding said grid workload for assignment to another of the at least two workstations.

In a computer network including at least two workstations and implementing a grid for executing a grid workload on the shared resources of the at least two workstations, a method of assigning said grid workload to one of the at least two workstations, the method comprising:

• running a hypervisor on each of the at least two workstations;
• executing a local workload on a first virtual machine running under each of said hypervisors, each hypervisor giving priority to said each first virtual machine;
• running at least one additional virtual machine under each hypervisor, said each at least one additional virtual machine executing an operating system capable of operating as part of said grid;
• monitoring a utilization of said each workstation by said each local workload and determining a local workload activity level indicative thereof;
• in response to a determination that said local workload activity level indicates said utilization is below a first threshold, sharing the resources of said workstation with said at least one additional virtual machine in order to execute said grid workload.

In a computer network including at least two workstations and implementing a grid for executing a grid workload on the shared resources of the at least two workstations, a system for assigning said grid workload to one of the at least two workstations, the system comprising:

• a workstation monitor for determining an activity state of a first workstation of the at least two workstations, said activity state being indicative of a degree of current utilization of said first workstation by a local workload;
• and
• a workload scheduler for;
• in response to a determination that said activity state indicates low utilization, assigning said grid workload to said first workstation;
• or
• in response to a determination that said activity state indicates high utilization, holding said grid workload for assignment to another of the at least two workstations.

In a computer network including at least two workstations and implementing a grid for executing a grid workload on the shared resources of the at least two workstations, a system for assigning said grid workload to one of the at least two workstations, the system comprising:

• a workstation monitor for:
• monitoring a utilization level indicative of the extent to which a local workload is utilizing the resources of a first workstation of the at least two workstations and setting a first workstation activity state to;
• idle if said utilization level is below a first threshold, or to
• active if said utilization level is above said first threshold and said first workstation activity state is already idle or active, or to
• blocked if said utilization level is above said first threshold and said first workstation activity state is already blocked, and;
• determining a first workstation preference for accepting said grid workload when said first workstation activity state is active;
• and p1 a workload scheduler for:
• checking said first workstation activity state;
• in response to a workstation activity state of idle, assigning said grid workload to said first workstation;
• in response to a first workstation activity state of active, checking said first workstation preference;
• in response to a first workstation preference of yes, assigning said grid workload to said first workstation;
• in response to a first workstation preferance of no;
• obtaining a second grid information for a group of machines including the
• setting said first workstation activity state to blocked;
• holding said grid workload for assignment to another
• downloading user interface logic data to the external processor when the of the at least two workstations;
• and
• in response to a first workstation activity state of blocked, holding said grid workload for assignment to another of the at least two workstations.

In a computer network including at least two workstations and implementing a grid for executing a grid workload on the shared resources of the at least two workstations, a system for assigning said grid workload to one of the at least two workstations, the system comprising:

• a hypervisor running on each of the at least two workstations;
• a local workload executing on a first virtual machine running under each of said hypervisors, said each hypervisor giving priority to said each first virtual machine;
• at least one additional virtual machine running under said each hypervisor, said each at least one additional virtual machine executing an operating system capable of operating as part of said grid;
• a workstation monitor for monitoring a utilization of said each workstation by said each local workload and determining a local workload activity level indicative thereof;
• a workload scheduler which, in response to a determination that said local workload activity level indicates said utilization is below a first threshold, shares the resources of said workstation with said at least one additional virtual machine in order to execute said grid workload.