On-demand compute environment

• US 9,112,813 B2
• Filed: 02/04/2013
• Issued: 08/18/2015
• Est. Priority Date: 03/16/2005
• Status: Active Grant

• Current Assignee: III Holdings 12, LLC (Intellectual Ventures LLC)
• Original Assignee: Adaptive Computing Enterprises, Inc. (Ala Services LLC)
• Inventors: Jackson, David Brian
• Primary Examiner(s): Eskandarnia, Arvin
• Assistant Examiner(s): DESAI, MARGISHI V
• Application Number:
  US13/758,164
• Publication Number:
  US 20130151711A1
• Time in Patent Office: 925 Days
• Field of Search: 709/225, 709/226
• US Class Current: 1/1
• CPC Class Codes: G06F 11/2038 with a single idle spare processing component
• CPC Class Codes: G06F 15/173 using an interconnection network, e.g. matrix, shuffle, pyramid, star, snowflake
• CPC Class Codes: G06F 2209/5014 Reservation
• CPC Class Codes: G06F 2209/509 Offload
• CPC Class Codes: G06F 9/5027 the resource being a machine, e.g. CPUs, Servers, Terminals
• CPC Class Codes: G06F 9/5061 Partitioning or combining of resources
• CPC Class Codes: G06F 9/5072 Grid computing
• CPC Class Codes: G06F 9/5077 Logical partitioning of resources; Management or configuration of virtualized resources specific details on emulation or internal functioning of virtual machines G06F9/455
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G08C 17/00 Arrangements for transmitting signals characterised by the use of a wireless electrical link
• CPC Class Codes: H04L 47/70 Admission control; Resource allocation
• CPC Class Codes: H04L 47/80 Actions related to the user profile or the type of traffic
• CPC Class Codes: H04L 47/83 based on usage prediction
• 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
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers
• CPC Class Codes: H04L 67/125 involving control of end-device applications over a network

Abstract

An on-demand compute environment comprises a plurality of nodes within an on-demand compute environment available for provisioning and a slave management module operating on a dedicated node within the on-demand compute environment, wherein upon instructions from a master management module at a local compute environment, the slave management module modifies at least one node of the plurality of nodes.

Claims

1. An on-demand compute environment comprising:

a processor; and

a computer-readable storage medium having stored therein instructions which, when executed by the processor, cause the processor to perform operations comprising;

receiving instructions from a master management module in an external compute environment separate from the on-demand compute environment to yield received instructions, wherein the received instructions comprise a specification of resources associated with a request for resources in the on-demand computer environment, the specification of resources being one of fully explicit, partially explicit, fully implicit based on workload, and based on a virtual private cluster package concept where the virtual private cluster package can include aspects of provisioning a support environment and adjustments to resource request timeframes including pre-allocation, allocation duration, and post-allocation timeframe adjustments, and wherein the external compute environment is configured to receive workload for processing on external compute nodes from users via reservation of the external compute nodes, and the on-demand compute environment has on-demand compute nodes available for provisioning and processing workload via reservation of the on-demand compute nodes under a slave management module;

provisioning a compute node in the on-demand compute environment based on the received instructions, the compute node to be managed by the slave management module to yield a provisioned compute node; and

receiving workload for processing from the external compute environment and causing the workload to consume the provisioned compute node, wherein the slave management module manages the provisioned compute node for the workload only for duration associated with how long the workload consumes the provisioned compute node.

2. The on-demand compute environment of claim 1, wherein the compute node is provisioned according to an analysis of at least part of the workload.

3. The on-demand compute environment of claim 1, wherein provisioning of the compute node further comprises modifying an existing reservation associated with the compute node.

4. The on-demand compute environment of claim 1, wherein the slave management module communicates with the master management module in the external compute environment.

5. The on-demand compute environment of claim 1, wherein provisioning the compute node further comprises provisioning a resource manager with the slave management module to manage the provisioned compute node.

6. A method of managing an on-demand compute environment, the method comprising:

receiving instructions at a slave management module, the slave management module operating on a dedicated node within the on-demand compute environment, from a master management module associated with an external compute environment separate from the on-demand compute environment to yield received instructions, wherein the master management module operates on a node within the external computing environment and wherein the received instructions comprise a specification of resources associated with a request for resources in the on-demand computer environment, the specification of resources being one of fully explicit, partially explicit, fully implicit based on workload, and based on a virtual private cluster package concept where the virtual private cluster package can include aspects of provisioning a support environment and adjustments to resource request timeframes including pre-allocation, allocation duration, and post-allocation timeframe adjustments, and wherein the external compute environment is configured to receive workload for processing on external compute nodes in the external compute environment from users via reservation of the external compute nodes, and the on-demand compute environment has on-demand compute nodes available for provisioning and processing workload via the slave management module via reserving the on-demand compute nodes;

provisioning, via the slave management module and based on the received instructions, on-demand compute nodes in the on-demand compute environment to yield provisioned on-demand compute nodes; and

receiving workload for processing from the external compute environment, wherein the slave management module manages the provisioned on-demand compute nodes for the workload only for duration of the workload consuming the provisioned on-demand compute nodes.

7. The method of claim 6, wherein the on-demand compute nodes are provisioned based on an analysis of at least part of the workload.

8. The method of claim 6, wherein the slave management module communicates with the master management module.

9. A non-transitory computer-readable storage medium having stored therein instructions which, when executed by a processor within an on-demand compute environment having on-demand compute nodes available for provisioning and for processing workload, cause the processor to perform operations comprising:

receiving instructions from a master management module in an external compute environment separate from the on-demand compute environment to yield received instructions, wherein the received instructions comprise a specification of resources associated with a request for resources in the on-demand computer environment, the specification of resources being one of fully explicit, partially explicit, fully implicit based on workload, and based on a virtual private cluster package concept where the virtual private cluster package can include aspects of provisioning a support environment and adjustments to resource request timeframes including pre-allocation, allocation duration, and post-allocation timeframe adjustments, and wherein the external compute environment is configured to receive workload for processing on external compute nodes in the external compute environment from users via registration of the external compute nodes, and the on-demand compute environment has on-demand compute nodes available for provisioning and processing workload via the slave management module via reserving the on-demand compute nodes;

provisioning, via the slave management module and based on the received instructions, an on-demand compute node to be managed by the slave management module to yield a provisioned compute node; and

receiving workload for processing from the external compute environment that consumes the provisioned on-demand compute node, wherein the slave management module operates to manage the provisioned compute node for the workload only while the workload consumes the provisioned on-demand compute node.

10. The non-transitory computer-readable storage medium of claim 9, wherein the on-demand compute node is provisioned according to an analysis of at least part of the workload.

11. The non-transitory computer-readable storage medium of claim 9, wherein provisioning of the on-demand other compute node further comprises modifying an existing reservation associated with the one on-demand compute node.

12. The non-transitory computer-readable storage medium of claim 9, wherein the slave management module communicates with the master management module in the external compute environment.

13. The non-transitory computer-readable storage medium of claim 9, wherein provisioning the compute node further comprises provisioning a resource manager with the slave management module to manage the provisioned on-demand compute node.

14. A method of managing access to a separately administered on-demand compute environment using a computing device in a compute environment having a master management module that manages compute resources in the compute environment, the method comprising:

identifying, via a processor, a triggering event in the compute environment, the compute environment configured to receive workload for processing via reserving compute nodes in the compute environment;

communicating via the master management module with a slave management module in the on-demand compute environment, the master management module operating on a node within the compute environment and the slave management module operating on a dedicated node within the on-demand compute environment;

transmitting instructions to the slave management module that cause the slave management module to provision on-demand compute nodes in the on-demand compute environment to be managed by the slave management module to yield provisioned on-demand compute nodes, wherein the instructions comprise a specification of resources associated with a request for resources in the on-demand computer environment, the specification of resources being one of fully explicit, partially explicit, fully implicit based on workload, and based on a virtual private cluster package concept where the virtual private cluster package can include aspects of provisioning a support environment and adjustments to resource request timeframes including pre-allocation, allocation duration, and post-allocation timeframe adjustments; and

communicating at least part of the workload to the on-demand compute environment, wherein the at least part of the workload consumes resources of the provisioned on-demand compute nodes.

15. The method of claim 14, wherein the slave management module reserves, allocates, and dynamically provisions compute resources associated with the workload in the on-demand compute environment.

16. The method of claim 14, further comprising transmitting to the on-demand compute environment second instructions for unprovisioning the provisioned on-demand compute nodes.

17. The method of claim 14, further comprising determining when to unprovision the provisioned on-demand compute nodes.

18. The method of claim 14, further comprising:

receiving updated instructions based on an update event associated with the workload for provisioning additional compute nodes in the on-demand compute environment to yield additional provisioned compute nodes; and

communicating part of the workload associated with the update event to at least one of the provisioned compute nodes and the additional provisioned compute nodes in the on-demand compute environment.