Method and apparatus for providing resource allocation policy

• US 8,850,449 B2
• Filed: 12/19/2008
• Issued: 09/30/2014
• Est. Priority Date: 12/19/2008
• Status: Active Grant

• Current Assignee: AT&T Intellectual Property I LP (AT&T, Inc.)
• Original Assignee: AT&T Intellectual Property I LP (AT&T, Inc.)
• Inventors: Hiltunen, Matti, Joshi, Kaustubh, Jung, Gueyoung, Pu, Calton, Schlichting, Richard
• Primary Examiner(s): Faherty, Corey S
• Application Number:
  US12/340,471
• Publication Number:
  US 20100162257A1
• Time in Patent Office: 2,111 Days
• Field of Search: None
• US Class Current: 718/108
• CPC Class Codes: G06F 9/50 Allocation of resources, e.g. of the central processing unit [CPU]

Abstract

A method and apparatus for providing a resource allocation policy in a network are disclosed. For example, the method constructs a queuing model for each application. The method defines a utility function for each application and for each transaction type of each application, and defines an overall utility in a system. The method performs an optimization to identify an optimal configuration that maximizes the overall utility for a given workload, and determines one or more adaptation policies for configuring the system in accordance with the optimal configuration.

Claims

1. A method for providing a resource allocation policy, comprising:

constructing a queuing model for each application;

defining a utility function for each application and for each transaction type of each application, wherein the transaction type is associated with a type of transaction performed by a corresponding application, wherein the transaction type comprises at least one of;

a blogging transaction, a profiling transaction, a browsing transaction, a logging-in transaction, a searching transaction, a buying transaction, and a selling transaction;

defining an overall utility in a system in accordance with the queuing model for each application, the utility function for each application and for each transaction type of each application;

performing an optimization to identify an optimal configuration that maximizes the overall utility for a given workload; and

determining an adaptation policy for configuring the system in accordance with the optimal configuration.

2. The method of claim 1, further comprising:

forward the adaptation policy to a rule-based management system.

3. The method of claim 1, further comprising:

verifying the adaptation policy.

4. The method of claim 1, further comprising:

notifying a user, a network administrator, or a system whether the adaptation policy is successfully implemented.

5. The method of claim 1, wherein the queuing model is a layered queuing network model.

6. The method of claim 5, wherein a parameter for the layered queuing network model is computed in an off-line training phase.

7. The method of claim 6, wherein the parameter is defined to represent:

a latency introduced by a network, a delay due to a disk input/output, a processing performed by a software component, a delay due to an interaction induced by a virtual machine environment, or a work load generation rate.

8. The method of claim 7, wherein the work load generation rate is a per transaction type.

9. The method of claim 1, wherein the performing the optimization comprises:

selecting an application configuration; and

determining an optimal component placement for the application configuration that is selected.

10. The method of claim 9, wherein the selecting the application configuration uses an optimization algorithm that is a discrete gradient-based search algorithm.

11. The method of claim 9, wherein the determining the optimal component placement is performed using a bin-packing algorithm.

12. The method of claim 11, wherein the bin-packing algorithm uses an algorithm based on a gradient-descent search for choosing a configuration change that yields a maximum reduction in overall central processing unit utilization for a unit reduction in utility.

13. The method of claim 11, wherein the bin-packing algorithm uses a central processing unit utilization as a volume for each node.

14. The method of claim 1, wherein the adaptation policy enables a selection of a replication level.

15. A non-transitory computer-readable medium having stored thereon a plurality of instructions, the plurality of instructions including instructions which, when executed by a processor, cause the processor to perform operations for providing a resource allocation policy, the operations comprising:

constructing a queuing model for each application;

defining a utility function for each application and for each transaction type of each application, wherein the transaction type is associated with a type of transaction performed by a corresponding application, wherein the transaction type comprises at least one of;

a blogging transaction, a profiling transaction, a browsing transaction, a logging-in transaction, a searching transaction, a buying transaction, and a selling transaction;

defining an overall utility in a system in accordance with the queuing model for each application, the utility function for each application and for each transaction type of each application;

performing an optimization to identify an optimal configuration that maximizes the overall utility for a given workload; and

determining an adaptation policy for configuring the system in accordance with the optimal configuration.

16. The non-transitory computer-readable medium of claim 15, further comprising:

forward the adaptation policy to a rule-based management system.

17. The non-transitory computer-readable medium of claim 15, further comprising:

verifying the adaptation policy.

18. The non-transitory computer-readable medium of claim 15, wherein the queuing model is a layered queuing network model.

19. The non-transitory computer-readable medium of claim 18, wherein the parameter for the layered queuing network model is computed in an off-line training phase.

20. An apparatus for providing a resource allocation policy, comprising:

a processor; and

a computer readable medium in communication with the processor, the computer readable medium has stored thereon instructions that, when executed by the processor, cause the processor to perform operations, the operations comprising;

constructing a queuing model for each application;

defining a utility function for each application and for each transaction type of each application, wherein each transaction type is associated with a type of transaction performed by a corresponding application, wherein the transaction type comprises at least one of;

a blogging transaction, a profiling transaction, a browsing transaction, a logging-in transaction, a searching transaction, a buying transaction, and a selling transaction;

defining an overall utility in a system in accordance with the queuing model for each application, the utility function for each application and for each transaction type of each application;

performing an optimization to identify an optimal configuration that maximizes the overall utility for a given workload; and

determining an adaptation policy for configuring the system in accordance with the optimal configuration.