SYSTEMS AND METHODS FOR GENERATING DYNAMICALLY CONFIGURABLE SUBSCRIPTION PARAMETERS FOR TEMPORARY MIGRATION OF PREDICTIVE USER WORKLOADS IN CLOUD NETWORK

• US 20120131594A1
• Filed: 11/24/2010
• Published: 05/24/2012
• Est. Priority Date: 11/24/2010
• Status: Active Grant

• Current Assignee: Red Hat, Inc. (International Business Machines Corporation)
• Original Assignee: Red Hat, Inc. (International Business Machines Corporation)
• Inventors: Morgan, Christopher Edwin
• Granted Patent: US 9,442,771 B2
• Time in Patent Office: Days
• US Class Current: 718/105
• CPC Class Codes: G06F 3/0647 Migration mechanisms
• CPC Class Codes: G06F 9/5072 Grid computing
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling

Abstract

Embodiments relate to systems and methods for generating dynamically configurable subscription parameters for the temporary migration of predictive user workloads in a cloud network. Aspects relate to platforms and techniques for analyzing overnight or other off-peak or temporary deployments of user workloads to underutilized host clouds. A cloud management system can capture usage history data for a user operating in a default deployment, such as a premise/cloud mix. A deployment engine can determine the resources required for the user'"'"'s workload pattern, and examine corresponding resources available in a set of other geographically-dispersed host clouds. The host clouds can comprise clouds based in different time zones, so that cloud capacity during U.S. West Coast evening time or European overnight hours can be packaged and offered to U.S. East Coast users at reduced rates. The deployment engine can generate different sets of dynamic subscription terms or parameters to be offered to the user, such as different costs or service levels at staggered off-peak periods.

Claims

1. A method of generating subscription parameters based on a predictive workload, comprising:

accessing a set of usage history data for a user operating a workload on a set of machines in a default deployment;

generating a predictive workload for the user based on the set of usage history data;

identifying a set of available resources in a set of host clouds provided by a cloud provider;

accessing a set of deployment criteria received from the cloud provider; and

generating a set of dynamically generated subscription parameters based on at least the predictive workload, the set of available resources in the set of host clouds, and the set of deployment criteria for migration of a workload of the user to the set of host clouds.

2. The method of claim 1, wherein the set of host clouds comprises a set of geographically dispersed host clouds.

3. The method of claim 2, wherein the set of geographically dispersed host clouds comprises at least one host cloud operating in a first UTC (coordinated universal time) time zone and a second host cloud operating in a second UTC (coordinated universal time) time zone.

4. The method of claim 3, wherein the predictive workload is predicted for the user in the second UTC (universal coordinated time) time zone for an off-peak time period for the user.

5. The method of claim 4, wherein the off-peak time period for the user is determined by analyzing the set of usage history data for the user.

6. The method of claim 5, wherein the set of deployment criteria comprises a least-cost criteria for the user based on migration of the predictive workload to the set of host clouds during the off-peak time period for the user.

7. The method of claim 6, wherein the set of dynamically generated subscription parameters comprises a reduced subscription cost based on execution of the predictive workload in the set of host clouds during the off-peak time period for the user.

8. The method of claim 7, wherein the set of dynamically generated subscription parameters comprises a service level agreement (SLA) to be maintained for host cloud resources in the set of host clouds during the off-peak time period for the user.

9. The method of claim 8, further comprising migrating the workload from the default deployment to the set of host clouds during the off-peak period for the user based on the set of dynamically generated subscription parameters.

10. The method of claim 9, further comprising migrating the workload from the set of host clouds to the default deployment after the off-peak period for the user.

11. The method of claim 4, wherein the off-peak time period for the user comprises a plurality of off-peak time periods for the user.

12. The method of claim 1, wherein the set of aggregate usage history data comprises at least one of processor usage data, memory usage data, storage usage data, communications bandwidth usage data, operating system usage data, application usage data, service usage data, virtual machine instance data, or appliance usage data.

13. A system for generating subscription parameters based on a predictive workload, comprising:

an interface to a data store, the data store storing a set of usage history data for a user operating a workload on a set of machines in a default deployment; and

a processor, communicating with the data store via the interface, the processor being configured to—

generate a predictive workload for the user based on the set of usage history data,

identify a set of available resources in a set of host clouds provided by a cloud provider;

access a set of deployment criteria received from the cloud provider, and

generate a set of dynamically generated subscription parameters based on at least the predictive workload, the set of available resources in the set of host clouds, and the set of deployment criteria for migration of a workload of the user to the set of host clouds.

14. The system of claim 13, wherein the set of host clouds comprises a set of geographically dispersed host clouds.

15. The system of claim 14, wherein the set of geographically dispersed host clouds comprises at least one host cloud operating in a first UTC (coordinated universal time) time zone and a second host cloud operating in a second UTC (coordinated universal time) time zone.

16. The system of claim 15, wherein the predictive workload is predicted for the user in the second UTC (universal coordinated time) time zone for an off-peak time period for the user.

17. The system of claim 16, wherein the off-peak time period for the user is determined by analyzing the set of usage history data for the user.

18. The system of claim 17, wherein the set of deployment criteria comprises a least-cost criteria for the user based on migration of the predictive workload to the set of host clouds during the off-peak time period for the user.

19. The system of claim 18, wherein the set of dynamically generated subscription parameters comprises a reduced subscription cost based on execution of the predictive workload in the set of host clouds during the off-peak time period for the user.

20. The system of claim 19, further comprising migrating the workload from the default deployment to the set of host clouds during the off-peak period for the user based on the set of dynamically generated subscription parameters.

21. The system of claim 20, further comprising migrating the workload from the set of host clouds to the default deployment after the off-peak period for the user.