Use of constraint-based linear programming to optimize hardware system usage

• US 8,418,184 B2
• Filed: 09/24/2010
• Issued: 04/09/2013
• Est. Priority Date: 09/24/2010
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Friedlander, Robert R., Hennessy, Richard, Kraemer, James R.
• Primary Examiner(s): ZHE, MENG YAO
• Application Number:
  US12/890,309
• Publication Number:
  US 20120079493A1
• Time in Patent Office: 928 Days
• Field of Search: 718/104
• US Class Current: 718/104
• CPC Class Codes: G06F 9/5027 the resource being a machine, e.g. CPUs, Servers, Terminals

Abstract

A computer implemented method, system, and/or computer program product optimizes systems usage. A work request is decomposed into units of work. A processor selectively sends each unit of work from the work request to either a first system or a second system for execution, depending on a work constraint on each unit of work and/or system constraints on the first and second systems.

Claims

1. A computer implemented method of optimizing systems usage, the computer implemented method comprising:

a processor receiving a work request;

the processor decomposing the work request into units of work;

the processor retrieving a first unit of work from the units of work;

the processor selectively sending the first unit of work to a first system or a second system for execution based on a first work constraint on the first unit of work, wherein the first work constraint limits execution of the first unit of work to a system that is compliant with an execution environment rule for the first unit of work; and

the processor minimizing a cost of running multiple units of work according to an algorithm;

2. The computer implemented method of claim 1, wherein the first system is a first local system, and wherein the second system is a first cloud system that provides on-demand network access to a shared pool of configurable computing resources.

3. The computer implemented method of claim 2, further comprising:

the processor sending the first unit of work to the first local system in response to determining that a transmission line to the first cloud system is overloaded.

4. The computer implemented method of claim 2, further comprising:

the processor sending the first unit of work to the first local system in response to determining that the first work constraint on the first unit of work is a prohibition against transmitting data related to the first unit of work over a transmission line to the first cloud system.

5. The computer implemented method of claim 2, further comprising:

the processor sending the first unit of work to the first cloud system in response to determining that the first work constraint on the first unit of work is a security constraint that prohibits the first unit of work from executing in the first local system.

6. The computer implemented method of claim 2, further comprising:

the processor sending the first unit of work to the first cloud system in response to determining that the first work constraint on the first unit of work is a legal constraint that prohibits the first unit of work from executing in the first local system.

7. The computer implemented method of claim 1, wherein the second system is a first cloud system to which the first unit of work was selectively sent, and wherein the computer implemented method further comprises:

the processor creating a second cloud system in response to determining that the first unit of work has not completed executing in the first cloud system within a predetermined amount of time; and

the processor sending the first unit of work to the second cloud system for execution.

8. The computer implemented method of claim 1, wherein the second system is a first cloud system to which the first unit of work was selectively sent, and wherein the computer implemented method further comprises:

the processor also sending the first unit of work to a second cloud system; and

the processor, in response to receiving a completed output of the first unit of work from the second cloud system without receiving the completed output of the first unit of work from the first cloud system, cancelling execution of the first unit of work in the first cloud system, wherein the first cloud system is executing the first unit of work at a slower rate than the second cloud system.

9. The computer implemented method of claim 1, further comprising:

the processor determining that the first work constraint on the first unit of work is a price budget; and

the processor selectively sending the first unit of work to a first local system or a first cloud system for execution depending on which system is cheaper.

10. The computer implemented method of claim 1, wherein the second system is a first cloud system to which the first unit of work was selectively sent, wherein resources in the first cloud system are physically located within a first time zone, and wherein the computer implemented method further comprises:

the processor identifying additional resources in a second time zone; and

the processor organizing the additional resources in the second time zone into a second cloud system for executing subsequent units of work from the work request.

11. The computer implemented method of claim 1, wherein multiple units of work are run on multiple cloud computer systems within a predetermined length of time, and wherein each unit of work from the multiple units of work run once and only once.

12. A computer program product for optimizing systems usage, the computer program product comprising:

a non-transitory computer readable storage media;

first program instructions to receive a work request;

second program instructions to decompose the work request into units of work;

third program instructions to retrieve a first unit of work from the units of work;

fourth program instructions to selectively send the first unit of work to a first system or a second system for execution based on a first work constraint on the first unit of work, wherein the first work constraint limits execution of the first unit of work to a system that is compliant with an execution environment rule for the first unit of work; and

fifth program instructions to minimize a cost of running multiple units of work according to an algorithm;

13. The computer program product of claim 12, wherein the first system is a first local system, and wherein the second system is a first cloud system that provides on-demand network access to a shared pool of configurable computing resources.

14. The computer program product of claim 13, further comprising:

sixth program instructions to send the first unit of work to the first local system in response to determining that a transmission line to the first cloud system is overloaded; and

wherein the sixth program instructions are stored on the non-transitory computer readable storage media.

15. The computer program product of claim 12, wherein the second system is a first cloud system to which the first unit of work was selectively sent, and wherein the computer program product further comprises:

sixth program instructions to create a second cloud system in response to determining that the first unit of work has not completed executing in the first cloud system within a predetermined amount of time; and

seventh program instructions to send the first unit of work to the second cloud system for execution;

and wherein the sixth and seventh program instructions are stored on the non-transitory computer readable storage media.

16. A computer system comprising:

a central processing unit (CPU), a computer readable memory, and a computer readable storage media;

first program instructions to receive a work request;

second program instructions to decompose the work request into units of work;

third program instructions to retrieve a first unit of work from the units of work;

fourth program instructions to selectively send the first unit of work to a first system or a second system for execution based on a first work constraint on the first unit of work, wherein the first work constraint limits execution of the first unit of work to a system that is compliant with an execution environment rule for the first unit of work; and

fifth program instructions to minimize a cost of running multiple units of work according to an algorithm;

17. The computer system of claim 16, wherein the first system is a first local system, and wherein the second system is a first cloud system that provides on-demand network access to a shared pool of configurable computing resources.

18. The computer system of claim 17, further comprising:

sixth program instructions to send the first unit of work to the first local system in response to determining that a transmission line to the first cloud system is overloaded;

and wherein the fifth program instructions are stored on the computer readable storage media for execution by the CPU via the computer readable memory.

19. The computer system of claim 17, wherein the second system is a first cloud system to which the first unit of work was selectively sent, and wherein the computer program product further comprises, and wherein the computer system further comprises:

sixth program instructions to retrieve a second unit of work from the units of work;

seventh program instructions to selectively send the second unit of work to the first cloud system or a second cloud system for execution depending on a second work constraint on the second unit of work; and

eighth program instructions to assemble work outputs from the first and second cloud systems for the first and second units of work into a unitary product; and

wherein the sixth, seventh, and eighth program instructions are stored on the computer readable storage media execution by the CPU via the computer readable memory.