Method of managing workloads and associated distributed processing system

• US 20050010664A1
• Filed: 07/06/2004
• Published: 01/13/2005
• Est. Priority Date: 03/30/2000
• Status: Active Grant

• Current Assignee: Byteweavr, LLC (Ascend Innovation Management, LLC)
• Original Assignee: United Devices, Inc. (Altair Engineering Incorporated)
• Inventors: Hubbard, Edward A.
• Granted Patent: US 7,092,985 B2
• Time in Patent Office: Days
• US Class Current: 709/224
• CPC Class Codes: G06F 2209/5021 Priority
• CPC Class Codes: G06F 9/5044 considering hardware capabilities

Abstract

A method for managing workloads and associated distributed processing system are disclosed that identify the capabilities of distributed devices connected together through a wide variety of communication systems and networks and utilize those capabilities to organize, manage and distribute project workloads to the distributed devices.

Claims

1-24. -24. (canceled)

25. A method operating a distributed processing system having a network coupling a multiplicity of Host distributed devices for processing workloads for the distributed processing system, a plurality of Client systems requesting processing of the workloads, a Server system for selectively distributing the workloads from the plurality of Client systems for processing by the distributed processing system, and a capability database coupled to the Server system comprising the steps of:

a) storing scheduling data corresponding to each of the plurality of Host distributed systems and to each of the plurality of Client systems users in the capability database;

b) generating a capability vector for each of the plurality of Host distributed devices with capability values defining an ability each of the plurality of Host distributed devices has for processing workloads;

c) generating a Client priority for each of the plurality of Client systems defining an order a particular Client system within the plurality of Client systems is granted access for sending a selected workload to the distributed processing system for processing, wherein the Client priority is stored as scheduling data in the capability database;

d) assigning a Client workload priority to each workload within each of the plurality of Client systems defining an order in which the selected workload is sent to the distributed processing system when the particular Client system is granted access to the distributed processing system in response to its corresponding Client priority, wherein the Client workload priorities are stored as scheduling data in the capability database;

e) distributing the selected workload from a particular Client system by the Server system to the first Host distributed device within the plurality of distributed

f) modifying the scheduling data in the capability database in response to receiving status inputs from the plurality of Host distributed devices and the plurality of Client systems.

26. The method of claim 25, wherein a status input is sent to the Server system from a Host distributed to modify the capability database in response to a change in availability of resources in the Host distributed device resulting from processing a workload.

27. The method of claim 25, wherein a status input is sent to the Server system from a Client system when the Client system changes a Client workload priority for any workload the Client system may submit to the Server system to distribute for processing on the distributed processing system.

28. The method of claim 25, wherein the scheduling data comprises the Client priority of the particular Client system, the Client workload priority of the selected workload within the particular Client system, and the capability values in the capability vector of the selected Host distributed device.

29. The method of claim 25 further comprising the steps of:

g) grouping selected Host distributed devices from the plurality of Host distributed devices to form one or more Host groups, wherein a particular Host group is formed in response to the fixed capability values in the capability vectors of the selected Host distributed devices; and

h) assigning a group value as a capability value to the capability vector of each Host distributed device when assigned to a Host group.

30. The method of claim 28, wherein the group values include shared resources available as long as the particular Host distributed device is assigned to a particular Host group.

31. The method of claim 30, wherein the Server system distributes a first workload from a selected Client system in response to the scheduling data and a request from the selected Client system to use the distributed processing system to process workload.

32. The method of claim 25, wherein the capability values for a Host distributed system comprises, fixed values defining operating performance of system components, data present at the Host distributed system for processing particular workloads, application software present at the Host distributed system for processing particular workloads, and group values limiting workload processing by the Host distributed to specific Client systems.

33. The method of claim 28, wherein selected of the plurality of Host distributed devices are grouped in a particular Host group in response to capability values in their corresponding capability vector and an incentive value.

34. The method of claim 33, wherein the incentive value defines a benefit the selected of the plurality of Host distributes devices realize by being assigned to a particular Host group.

35. The method of claim 34, wherein the Server system generates incentive values available to the plurality of Host distributed devices for accepting an assignment to a particular Host group.

36. The method of claim 28, wherein the group capability values include an affiliation indication.

37. The method of claim 36, wherein the affiliation indication comprises an association with a particular educational institution or a particular scientific research project.

38. The method of claim 36, wherein the group capability values include a plurality of project types.

39. The method of claim 38 further comprising the step of allocating percentages of the processing capability of a Host distributed device grouped within a particular Host group between selected of the plurality of project types.

40. The method of claim 25, wherein a value of the Client priority of a particular Client system is set in response to a payment paid by the Client system for access to the distributed processing system.

41. The method of claim 28, wherein a value of the Client priority of a particular Client system is set in response to the Client system'"'"'s association with a formed Host group.

50. The computer program product of claim 28, wherein selected of the plurality of Host distributed devices are grouped in a particular Host group in response to capability values in their corresponding capability vector and an incentive value.

51. The computer program product of claim 50, wherein the incentive value defines a benefit the selected of the plurality of Host distributes devices realize by being assigned to a particular Host group.

52. The computer program product of claim 51, wherein the Server system generates incentive values available to the plurality of Host distributed devices for accepting an assignment to a particular Host group.

42. A computer program product operating within a Server system coupled to a network and managing a distributed processing system, the network configured to enable the Server system to selectively couple a multiplicity of Host distributed devices to perform workloads for the distributed processing system, wherein the server system is coupled to a capability database, the program product comprising a program of instructions for performing the program steps of:

a) storing scheduling data corresponding to each of the plurality of Host distributed systems and to each of the plurality of Client systems users in the capability database;

b) generating a capability vector for each of the plurality of Host distributed devices with capability values defining an ability each of the plurality of Host distributed devices has for processing workloads;

c) generating a Client priority for each of the plurality of Client systems defining an order a particular Client system within the plurality of Client systems is granted access for sending a selected workload to the distributed processing system for processing;

d) assigning a Client workload priority to each workload within each of the plurality of Client systems defining an order in each the selected workload is sent to the distributed processing system when the particular Client system is granted access to the distributed processing system in response to its corresponding Client priority;

e) distributing the selected workload from a particular Client system by the Server system to the first Host distributed device within the plurality of distributed devices in response to the scheduling data stored in the capability database; and

f) modifying the scheduling data in the capability database in response to receiving status inputs from the plurality of Host distributed devices and the plurality of Client systems.

43. The computer program product of claim 42, wherein a status input is sent to the Server system from a Host distributed to modify the capability database in response to a change in availability of resources in the Host distributed device resulting from processing a workload.

44. The computer program product of claim 42, wherein a status input is sent to the Server system from a Client system when the Client system changes a Client workload priority for any workload the Client system may submit to the Server system to distribute for processing on the distributed processing system.

45. The computer program product of claim 42, wherein the scheduling data comprises the Client priority of the particular Client system, the Client workload priority of the selected workload within the particular Client system, and the capability values in the capability vector of the selected Host distributed device.

46. The computer program product of claim 42 further comprising the steps of:

g) grouping selected Host distributed devices from the plurality of Host distributed devices to form one or more Host groups, wherein a particular Host group is formed in response to the fixed capability values in the capability vectors of the selected Host distributed devices; and

h) assigning a group value as a capability value to the capability vector of each Host distributed device when assigned to a Host group.

47. The computer program product of claim 45, wherein the group values include shared resources available as long as the particular Host distributed device is assigned to a particular Host group.

48. The computer program product of claim 47, wherein the Server system distributes a first workload from a selected Client system in response to the scheduling data and a request from the selected Client system to use the distributed processing system to process workload.

49. The computer program product of claim 42, wherein the capability values for a Host distributed system comprises, fixed values defining operating performance of system components, data present at the Host distributed system for processing particular workloads, application software present at the Host distributed system for processing particular workloads, and group values limiting workload processing by the Host distributed to specific Client systems.

53. A software agent operating within one of a multiplicity of Host distributed devices coupled to a network, the network configured to enable a Server system to selectively couple the multiplicity of Host distributed devices to perform workloads for the distributed processing system, the software agent comprising a program of instructions for performing the program steps of:

monitoring resource utilization within a Host distributed device selected from the multiplicity of Host distributed devices;

generating a status for the Host distributed device comprising types and capacities of resources available in the Host distributed system and utilizations of resources within the Host distributed device;

modifying the status for the Host distributed system whenever resources are added or deleted or when the availability of the resources change, thereby generating a status change; and

sending the status of the Host distributed device to the Server system in response to a status request or automatically in response to the status change.