SYSTEMS AND METHODS FOR SCHEDULING OF WORKLOAD-AWARE JOBS ON MULTI-CLOUDS

• US 20170357532A1
• Filed: 06/12/2017
• Published: 12/14/2017
• Est. Priority Date: 06/10/2016
• Status: Active Grant

• Current Assignee: Board of Regents of the University of Texas System (University of Texas System)
• Original Assignee: Board of Regents of the University of Texas System (University of Texas System)
• Inventors: Miraftabzadeh, Seyed Ali, Najafirad, Peyman
• Granted Patent: US 10,452,451 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 9/505 considering the load
• CPC Class Codes: G06F 9/5083 Techniques for rebalancing the load in a distributed system
• CPC Class Codes: G06Q 30/0283 Price estimation or determination
• CPC Class Codes: H04L 41/0895 Configuration of virtualised networks or elements, e.g. virtualised network function or OpenFlow elements
• CPC Class Codes: H04L 41/0896 Bandwidth or capacity management, i.e. automatically increasing or decreasing capacities flow or congestion control using dynamic resource allocation, e.g. in-call renegotiation, H04L47/76
• CPC Class Codes: H04L 41/40 using virtualisation of network functions or resources, e.g. SDN or NFV entities
• CPC Class Codes: H04L 41/5019 Ensuring fulfilment of SLA
• CPC Class Codes: H04L 41/5096 wherein the managed service relates to distributed or central networked applications
• CPC Class Codes: H04L 67/1008 based on parameters of servers, e.g. available memory or workload monitoring of computer activity G06F11/30

Abstract

Disclosed are systems and methods for optimally provisioning tasks to machine instances. The machine instances can correspond to cloud providers and can be part of a multi-cloud. The machine instances can each have various resources available. The task can be received from a user including a job profile. A ranking can be determined for the machine instances. Based on the ranking, the task can be provisioned to one or more of the machine instances.

Claims

1. A system comprising:

a plurality of machine instances individually corresponding to a respective amount of available resources;

a core system comprising at least one computing device, the core system configured to at least;

receive at least one task from a user, the at least one task comprising a job profile;

determining a rank for the plurality of machine instances based at least in part on the respective amount of available resources for each of the plurality of machine instances;

determine an identified machine instance of the plurality of machine instances based at least in part on the rank; and

provision the identified machine instance to perform the at least one task.

2. The system of claim 1, wherein subsets of the plurality of machine instances individually correspond to a plurality of cloud providers.

3. The system of claim 2, wherein the core system is further configured to at least monitor the plurality of cloud providers, and update a parameter associated with one of the plurality of cloud providers.

4. The system of claim 1, wherein the core system is further configured to at least monitor a work-load associated with the at least one task.

5. The system of claim 4, wherein the core system is further configured to at least update a parameter in the job profile based at least in part on monitoring the work-load.

6. The system of claim 1, wherein the job profile comprises at least one sigmoidal model.

7. The system of claim 1, wherein the job profile comprises at least one logarithmic model.

8. The system of claim 1, wherein a subset of the plurality of machine instances comprises a multi-cloud.

9. The system of claim 1, wherein the identified machine instance is determined based at least in part on whether the identified machine instance can perform the at least one task while satisfying requirements of a service level agreement (SLA).

10. The system of claim 1, wherein the core system is further configured to at least generate a plurality of clutches, each of the plurality of clutches including at least two instances from the plurality of instances.

11. A method comprising:

observing at least one dynamic parameter and at least one static parameter;

generating weights for a plurality of machine instances based at least in part on the at least one dynamic parameter and the at least one static parameter;

generating a ranking the plurality of machine instances and a plurality of clutches based at least in part on the weights;

determining an identified machine instance of the plurality of machine instances based at least in part on the ranking and the plurality of clutches; and

provisioning the identified machine instance to perform at least one task.

12. The method of claim 11, further comprising receiving a request to perform the at least one task from a user.

13. The method of claim 11, further comprising determining whether the identified machine instance can perform the at least one task while satisfying a plurality of constraints indicated in a service level agreement (SLA).

14. The method of claim 13, wherein determining whether the identified machine instance can perform the at least one task while satisfying the plurality of constraints is based at least in part on a sigmoidal utility function and a logarithmic utility function.

15. The method of claim 11, wherein the at least one task comprises a sigmoidal-like task associated with metadata defining an amount of resources requested for the task.

16. The method of claim 11, wherein the at least one task comprises a logarithmic task configured to use all available resources in a pool of the plurality of machine instances.

17. The method of claim 16, wherein the logarithmic task is configured to require a minimum quantity of resources.

18. The method of claim 16, wherein the logarithmic task corresponds to a logarithmic utility function represented by:

19. The method of claim 11, further comprising determining a dark price for use of the identified machine instance based at least in part on a dual problem.

20. The method of claim 19, generating a bid by weighing the dark price by the ranking of the plurality of machine instances.