Systems and methods to allocate application tasks to a pool of processing machines

• US 8,429,660 B2
• Filed: 08/22/2005
• Issued: 04/23/2013
• Est. Priority Date: 08/23/2004
• Status: Active Grant

• Current Assignee: Goldman Sachs & Co. LLC (Goldman Sachs Group Incorporated)
• Original Assignee: Goldman Sachs & Company (Goldman Sachs Group Incorporated)
• Inventors: Dweck, Jay S., Fellows, Michael, Richard, David G., Morrissey, Mark T., Byron, Mary D
• Primary Examiner(s): Kawsar, Abdullah Al
• Application Number:
  US11/208,991
• Publication Number:
  US 20060070078A1
• Time in Patent Office: 2,801 Days
• Field of Search: 718/107, 718 1-106, 705/5, 705/80, 705 51- 53, 705 722- 726, 709/201, 709/202, 709/203, 709220-226
• US Class Current: 718/104
• CPC Class Codes: G06F 2209/5011 Pool
• CPC Class Codes: G06F 2209/5014 Reservation
• CPC Class Codes: G06F 2209/5017 Task decomposition
• CPC Class Codes: G06F 9/50 Allocation of resources, e.g. of the central processing unit [CPU]
• CPC Class Codes: G06F 9/5027 the resource being a machine, e.g. CPUs, Servers, Terminals
• CPC Class Codes: H04L 67/10 in which an application is distributed across nodes in the network software deployment G06F8/60; multiprogramming arrangements G06F9/46

Abstract

Systems and methods are provided to allocate application tasks to a pool of processing machines. According to some embodiments, a requestor generates a scope request including an indication of a number of compute units to be reserved. The requestor also provides an application request associated with the scope. A subset of available processing machines may then be allocated to the scope, and the application request is divided into a number of different tasks. Each task may then be assigned to a processing machine that has been allocated to the application request. According to some embodiments, each task is associated with a deadline. Moreover, according to some embodiments an overall cost is determined and then allocated to the requestor based on the number of compute units that were reserved for the scope.

Claims

1. A processor-implemented method, comprising:

receiving, via processor, from a requestor a scope request including an indication of a number of compute units to be reserved for a scope, the compute units representing a quantifiable parameter associated with processing machines that define an amount of processing power;

receiving, via the processor, an application request associated with the scope, wherein the scope request includes an indication of a subscription period;

determining, via the processor, a requested number of compute units for the application request;

allocating, via the processor, the requested number of compute units and an additional number of compute units for the application request, wherein allocating comprises allocating at least one of any available processing machines from a pool of processing machines having the requested number of compute units and the additional number of compute units to the application request, and wherein other processing machines from the pool are allocated to other application requests associated with other scopes;

dividing, via the processor, the application request into a plurality of tasks;

associating, via the processor, a deadline with each task;

assigning, via the processor, each task to one of the allocated processing machines;

if a processing machine does not complete the task before the deadline, reassigning, via the processor, the task to another allocated processing machine; and

allocating, via the processor, a cost to the requestor, wherein the allocated cost is based at least in part on a number of compute units reserved for the scope and a cost-per-compute unit computed by dividing an overall cost associated with the entire pool of processing machines by a total number of compute units representing the processing power associated with all of the processing machines.

2. The method of claim 1, further comprising determining a number of processing machines to be provided in the entire pool of processing machines based on scope requests from a plurality of requestors.

3. The method of claim 1, wherein the cost allocated to the requestor is further based on the subscription period.

4. The method of claim 1, wherein the scope request further includes an indication of a time period associated with compute units to be reserved for the scope.

5. The method of claim 4, wherein the time period is associated with at least one of:

(i) a time of day, (ii) a time of month, (iii) a time of year.

6. The method of claim 1, wherein the number of compute units reserved for the scope represents a maximum number of compute units that can be assigned to the scope at one time.

7. The method of claim 6, further comprising:

determining if a total number of compute units currently assigned to the scope exceeds the number of compute units reserved for the scope.

8. The method of claim 1, wherein the number of compute units reserved for the scope represents a maximum number of compute unit time increments that can be assigned to the scope over a time period.

9. The method of claim 8, further comprising:

determining if a total number of compute unit time increments that have been used by the scope exceeds the number of compute unit time units reserved for the scope.

10. The method of claim 1, wherein said dividing is performed by at least one of:

(i) a programmer, (ii) an administrator, and (in) an automated scheduling/distribution manager.

11. The method of claim 1, wherein the overall cost associated with the entire pool of processing machines includes at least one of:

(i) a capital cost, (ii) an occupancy cost, (iii) a power cost, (iv) a cooling cost, and (v) an administrative cost.

12. The method of claim 1, wherein the application request is associated with a financial application.

13. The method of claim 1, wherein the requestor is remote from the processing machines.

14. The method of claim 1, where the processing machines comprise boards in a processing cabinet.

15. The method of claim 1, wherein the application is at least one of (i) resilient, and (ii) free of dependencies on a fixed machine address.

16. A non-transitory computer-readable medium storing instructions that when executed by a processor performs a method, said method comprising:

receiving from a requestor a scope request including an indication of a number of compute units to be reserved for the scope, the compute units representing a quantifiable parameter associated with processing machines that define an amount of .processing power;

receiving an application request associated with the scope, wherein the scope request includes an indication of a subscription period;

determining a requested number of compute units for the application request;

allocating, via the processor, the requested number of compute units and an additional number of compute units for the application request, wherein allocating comprises allocating at least one of any available processing machines from a pool of processing machines having the requested number of compute units and the additional number of compute units to the application request, and wherein other processing machines from the pool are allocated to other application requests associated with other scopes;

dividing, via the processor, the application request into a plurality of tasks;

associating, via the processor, a deadline with each task;

assigning, via the processor, each task to one of the allocated processing machines;

if a processing machine does not complete the task before the deadline, reassigning, via the processor, the task to another allocated processing machine; and

allocating a cost to the requestor, wherein the allocated cost is based at least in part on a number of compute units reserved for the scope and a cost-per-compute unit computed by dividing an overall cost associated with the entire pool of processing machines by a total number of compute units representing processing power associated with all of the processing machines.

17. The medium of claim 16, wherein the method further comprises:

determining processing machines to be provided in the entire pool of processing machines based on scope requests from a plurality of requesters.

18. The medium of claim 16, wherein the cost allocated to the requester is further based on the subscription period.

19. The medium of claim 16, wherein the overall cost associated with the entire pool of processing machines includes at least one of:

(i) a capital cost, (ii) an occupancy cost, (iii) a power cost, (iv) a cooling cost, and (v) an administrative cost.

20. The medium of claim 16, wherein the application request is associated with a financial transaction.

21. The medium of claim 16, wherein the requester is remote from the processing machines.

22. A scheduling/distribution manager, comprising:

a processor; and

a storage device in communication with said processor and storing instructions that when executed by said processor, causes the processor to;

receive from a requestor a scope request including an indication of a number of compute units to be reserved for a scope, the compute units representing a quantifiable parameter associated with processing machines that define an amount of processing power for the processing machines based on a number of processors, processor speed, an amount of memory, and hard disk drive storage;

receive an application request associated with the scope, wherein the scope request includes an indication of a subscription period;

determine a requested number of compute units for the application request;

allocate the requested number of compute units and an additional number of compute units for the application request, wherein allocating comprises allocating at least one of any available processing machines from a pool of processing machines having the requested number of compute units and the additional number of compute units to the application request, and wherein other processing machines from the pool are allocated to other application requests associated with other scopes;

divide the application request into a plurality of tasks;

associate a deadline with each task;

assign each task to one of the allocated processing machines;

if a processing machine does not complete the task before the deadline, reassign the task to another allocated processing machine; and

allocate a cost to the requestor, wherein the allocated cost is based at least in part on a number of compute units reserved for the scope and a cost-per-compute unit computed by dividing an overall cost associated with the entire pool of processing machines by a total number of compute units representing the processing power associated with all of the processing machines.

23. A system, comprising:

a processing machine pool; and

a scheduling/distribution manager, comprising;

a processor; and

a storage device in communication with said processor and storing instructions when executed by said processor, causes the processor to;

receive from a requestor a scope request including an indication of a number of compute units reserved for a scope, the compute units representing a quantifiable parameter associated with processing machines that define an amount of processing power;

receive an application request associated with the scope, wherein the scope request includes an indication of a subscription period;

determine a requested number of compute units for the application request;

allocate the requested number of compute units and an additional number of compute units for the application request, wherein allocating comprises allocating at least one of any available processing machines from a pool of processing machines having the requested number of compute units and the additional number of compute units to the application request, and wherein other processing machines from the pool are allocated to other application requests associated with other scopes;

divide the application request into a plurality of tasks;

associate a deadline with each task;

assign each task to one of the allocated processing machines;

if a processing machine does not complete the task before the deadline, reassign the task to another allocated processing machine; and

allocate a cost to the requestor, wherein the allocated cost is based at least in part on a number of compute units reserved for the scope and a cost-per-compute unit computed by dividing an overall cost associated with the entire pool of processing machines by a total number of compute units representing the processing power associated with all of the processing machines.