Self-adjusting framework for managing device capacity

Patent metadata

Field	Value
Publication number	US-9292354-B2
Application number	US-201314057753-A
Country	US
Kind code	B2
Filing date	Oct 18, 2013
Priority date	Oct 18, 2013
Publication date	Mar 22, 2016
Grant date	Mar 22, 2016

Abstract

Automatically improving a deployment. A method includes, in a live distributed computing environment, adjusting operating parameters of deployment components. Effects of the adjusted operating parameters are observed. At least a portion of a behavior model function is defined based on the adjusted operating parameters and observed effects. Based on current distributed computing environmental conditions, operating parameters defined in the behavior model function are adjusted to improve the deployment.

First claim

What is claimed is:

1. In a distributed computing environment for a distributed service or application, a method of automatically improving a deployment, the method comprising:
in a live distributed computing environment, adjusting operating parameters of deployment components;
observing effects of the adjusted operating parameters;
defining at least a portion of a behavior model function based on the adjusted operating parameters and observed effects; and
based on current distributed computing environmental conditions, adjusting operational parameters defined in the behavior model function to improve the deployment wherein adjusting the operational parameters is performed on an experimental parallel component implemented in parallel with a working component, while not adjusting parameters on the working component, the method further comprising:
determining that a predetermined confidence level has been achieved in the experimental component; and
as a result of determining that a predetermined confidence level has been achieved in the experimental component, applying the adjusted parameters of the experimental component to the working component.

2. The method of claim 1 , further comprising:
identifying how a load is expected to change over time; and
adjusting operational parameters defined in the behavior model function in anticipation of the load changing over time.

3. The method of claim 2 , further comprising:
determining how long applied adjustments to operational parameters take to become effective; and
wherein adjusting operational parameters defined in the behavior model function in anticipation of the load changing over time is performed in view of how long applied adjustments to operational parameters take to become effective.

4. The method of claim 1 , wherein the behavior model function comprises one or more parameters related to rates charged for components.

5. The method of claim 1 , wherein the behavior model function comprises one or more parameters related to geometry of components in the live distributed computing environment.

6. The method of claim 1 , wherein the behavior model function comprises one or more parameters related to configuration settings of components in the live distributed computing environment.

7. The method of claim 1 , wherein the behavior model function comprises one or more parameters related to topology of the live distributed computing environment.

8. The method of claim 1 , wherein the behavior model function comprises one or more parameters related to quality of service (QoS) required for the live distributed computing environment.

9. The method of claim 1 , wherein adjusting operating parameters defined in the behavior model function is performed until a particular pre-defined threshold is reached.

10. The method of claim 1 , wherein adjusting operating parameters defined in the behavior model function comprises adjusting a plurality of different parameters in parallel.

11. The method of claim 1 , further comprising using the adjusted operating parameters to start a new instance of a service.

12. The method of claim 1 , wherein the behavior model function is a function of an entire distributed service.

13. The method of claim 1 , wherein adjusting operating parameters defined in the behavior model function is performed by suspending activities based on the activities being lower priority than other activities.

14. In a distributed computing environment for a distributed service or application, a method of automatically improving a deployment based on anticipated loads to the deployment, the method comprising:
in a live distributed computing environment, identifying how a load is expected to change over time;
determining how long applied adjustments to operational parameters of deployment components in the live distributed computing environment take to become effective; and
adjusting operational parameters of deployment components in anticipation of the load changing over time, wherein adjusting operational parameters is performed on an experimental parallel component implemented in parallel with a working component, while not adjusting parameters on the working component, the method further comprising:
determining that a predetermined confidence level has been achieved in the experimental component; and

as a result of determining that a predetermined confidence level has been achieved in the experimental component, applying the adjusted parameters of the experimental component to the working component.

15. The method of claim 14 , wherein adjusting operational parameters of deployment components is performed by adjusting operating parameters defined in a behavior model function.

16. In a distributed computing environment for a distributed service or application, a system for automatically improving a deployment, the system comprising:
one or more processors; and
one or more computer readable media, wherein the one or more computer readable media comprise computer executable instructions that when executed by at least one of the one or more processors cause the system to perform the following:
in a live distributed computing environment, adjusting operating parameters of deployment components;
observing effects of the adjusted operating parameters;
defining at least a portion of a behavior model function based on the adjusted operating parameters and observed effects; and
based on current distributed computing environmental conditions, adjusting operational parameters defined in the behavior model function to improve the deployment, wherein adjusting the operational parameters is performed on an experimental parallel component implemented in parallel with a working component, while not adjusting parameters on the working component, the method further comprising:
determining that a predetermined confidence level has been achieved in the experimental component; and
as a result of determining that a predetermined confidence level has been achieved in the experimental component, applying the adjusted parameters of the experimental component to the working component.

17. The system of claim 16 , further comprising computer executable instructions that when executed by at least one of the one or more processors cause the system to perform the following:
identifying how a load is expected to change over time; and
adjusting operational parameters defined in the behavior model function in anticipation of the load changing over time.

18. The system of claim 17 , further comprising computer executable instructions that when executed by at least one of the one or more processors cause the system to perform the following:
determining how long applied adjustments to operational parameters take to become effective; and
wherein adjusting operational parameters defined in the behavior model function in anticipation of the load changing over time is performed in view of how long applied adjustments to operational parameters take to become effective.

19. The system of claim 17 , wherein the observed effects include at least one of a response time or an error rate.

20. The system of claim 16 , wherein adjusting operating parameters defined in the behavior model function is performed until a particular pre-defined threshold is reached.

Assignees

• Microsoft Corp
• Microsoft Technology Licensing Llc

Inventors

• Michael Michael Leonida
• Sanghvi Ashvinkumar Jayantilal
• Onalan Bahadir Baris
• Kapila Gaurav

Classifications

• G06F9/5077 Primary

  Logical partitioning of resources; Management or configuration of virtualized resources (specific details on emulation or internal functioning of virtual machines G06F9/455) · CPC title

• G06F9/5072 Primary

  Grid computing · CPC title

• G06N5/04

  Inference or reasoning models · CPC title