Adaptive upgrade to computing systems

• US 9,886,262 B2
• Filed: 02/24/2016
• Issued: 02/06/2018
• Est. Priority Date: 03/16/2015
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Thomas, Satish J., Kaminski, Tomasz K., Jewsbury, Andrew William
• Primary Examiner(s): Zhen, Wei
• Assistant Examiner(s): Luu, Binh
• Application Number:
  US15/052,058
• Publication Number:
  US 20160274893A1
• Time in Patent Office: 713 Days
• Field of Search: None
• CPC Class Codes: G06F 8/61 Installation
• CPC Class Codes: G06F 8/65 Updates security arrangements therefor G06F21/57
• CPC Class Codes: G06F 8/71 Version control security arrangements therefor G06F21/57; Configuration management
• CPC Class Codes: G06F 9/44505 Configuring for program initiating, e.g. using registry, configuration files
• CPC Class Codes: G06F 9/44536 Selecting among different versions

Abstract

An upgrade or update that is available to a deployed computing system (or deployment) under analysis is identified. Performance improvement metrics measured at similar deployments, that incorporated the identified update, are accessed to estimate a performance improvement metric corresponding to incorporating the identified update at the deployment under analysis.

Claims

1. A computing system, comprising:

a process modeling component configured to;

identify a computing process deployed on a target computing system;

an applicable upgrade identification system configured to;

based on the computing process, identify an upgrade applicable to the target computing system;

a machine learning system configured to;

identify another computing system as being similar to the target computing system based on

identification of the computing process deployed in the other computing system, and

a usage pattern indicative of usage of the computing process in the target computing system relative to a usage pattern indicative of usage of the computing process in the other computing system; and

receive a measured performance metric corresponding to the other computing system configured with the upgrade; and

based on the measured performance metric corresponding to the other computing system, generate a performance change metric indicative of an estimated change in a performance metric for the target computing system corresponding to incorporation of the upgrade into the target computing system; and

an upgrade system configured to;

select the upgrade based on the performance change metric; and

apply the upgrade to the target computing system.

2. The computing system of claim 1 wherein the update system is configured to:

measure the change in the performance metric based on the application of the upgrade; and

update the machine learning system based on the measured change in the performance metric.

3. The computing system of claim 1 wherein the applicable upgrade identification system comprises:

upgrade identifier logic configured to identify, as applicable upgrades, available upgrades to the deployed processes in the set of deployed ,processes; and

weight generation logic configured to generate an impact weight indicative of an impact of each of the applicable upgrades on the deployed processes.

4. The computing system of claim 3 wherein the process modeling component is further configured to:

identify usage patterns corresponding to the set of deployed processes indicative of usage of the set of deployed processes in the target computing system, the weight generation logic configured to generate the impact weight as having more impact if a given available upgrade affects a deployed process that is heavily used in the target computing system based on the identified usage patterns.

5. The computing system of claim 4 wherein the machine learning system comprises:

an upgrade effort estimation component configured to;

access a measured effort metric corresponding to the other computing system, the measured effort metric being indicative of effort used in configuring the other computing system with the applicable upgrade, and

based on the measured effort metric, generate an upgrade effort metric indicative of an estimated, measurable effort for configuring the target computing system with the applicable upgrade.

6. The computing system of claim 5 wherein the recommendation engine comprises:

recommended upgrade identifier logic configured to compare the upgrade effort metric to the performance change metric to identify a recommended upgrade to be included in the upgrade recommendation.

7. The computing system of claim 1 wherein the machine learning system is configured to:

obtain an upgrade criterion associated with the target computing system; and

generate the performance change metric relative to the upgrade criterion.

8. The computing system of claim 7 wherein the process modeling component is configured to identify a customization to the identified computing process in the target computing system.

9. The computing system of claim 8 wherein the machine learning system is configured to identify the other computing system based on

identification of the customization to the computing process in the other computing system, and

the upgrade criterion for the target computing system relative to an upgrade criterion associated with the other computing system.

10. The computing system of claim 9 wherein the recommendation engine comprises:

sequence identifier logic configured to;

based on a determination that the upgrade recommendation identifies a plurality of recommended upgrades, identify a recommended upgrade sequence for configuring the target computing system with the plurality of recommended upgrades.

11. The computing system of claim 10 and further comprising:

an upgrade component configured to;

automatically provision a test environment deploy the recommended upgrades into the test environment; and

measure the performance change metric based on configuring the target computing system with the recommended upgrades in the test environment,

the recommendation engine being configured to update the recommendation, based on the measured performance change metric, measured in the test environment.

12. The computing system of claim 11 wherein the upgrade component is further configured to:

perform the recommended upgrades in a production environment corresponding to the target computing system;

measure the effort metric corresponding to performing the recommended upgrades in the production environment;

measure the change in the performance metric based on performing the recommended upgrades in the production environment; and

provide, to the machine learning system, the measured effort metric corresponding to performing the recommended upgrades in the production environment and the measured change in the performance metric.

13. A computer implemented method, comprising:

identifying a computing process deployed on a target computing system;

based on the computing process, identifying an upgrade that is applicable to the target computing system;

identifying another computing system as being similar to the target computing system based on

identification of the computing process deployed in other computing system, and

a usage pattern indicative of usage of the computing process in the target computing system relative to a usage pattern indicative of usage of the computing process in the other computing system; and

receiving a measured performance metric corresponding to the other computing system configured with the upgrade;

based on the measured performance metric corresponding to the other computing system, generating a performance change metric indicative of an estimated change in a performance metric for the target computing system corresponding to incorporation of the upgrade into the target computing system;

selecting the upgrade based on the performance change metric; and

applying the upgrade to the target computing system.

14. The computer implemented method of claim 13 and further comprising:

generating an impact weight indicative of an impact of the upgrade on the deployed computing process.

15. The computer implemented method of claim 14 wherein generating an impact weight comprises:

generating the impact weight as having more impact if the upgrade affects a deployed computing process that is heavily used in the target computing system based on the usage pattern.

16. The computer implemented method of claim 15 and further comprising:

obtaining a measured effort metric corresponding to the other computing system, the measured effort metric being indicative of effort used in configuring the other computing system with the applicable upgrade; and

generating an upgrade effort metric indicative of an estimated, measurable effort for incorporating the applicable upgrade into the target computing system.

17. The computer implemented method of claim 16 wherein identifying a computing process comprises:

identifying a customization to the computing process in the target computing system.

18. The computer implemented method of claim 17 wherein the other computing system is identified as similar to the target computing system based on

identification of the customization to the computing process in the other computing system.

19. A computing system, comprising:

a process modeling component configured to;

identify a computing process deployed on a target computing system;

identify a usage pattern corresponding to the computing process indicative of usage of the computing process in the target computing system; and

identify a customization to the computing process in the target computing system;

an applicable upgrade identification system configured to;

based on the computing process, identify an upgrade applicable to the target computing system;

a machine learning system configured to;

identify another computing system as being similar to the target computing system based on

identification of the computing process deployed in other computing system, and

a usage pattern indicative of usage of the computing process in the target computing system relative to a usage pattern indicative of usage of the computing process in the other computing system; and

receive a measured performance metric corresponding to the other computing system configured with the upgrade; and

based on the measured performance metric correspond the other computing system, generate a performance charge metric indicative of an estimated change in a performance metric for the target computing system corresponding to incorporation of the upgrade into the target computing system; and

an upgrade system configured to;

select the upgrade based on the performance change metric;

apply the upgrade to the target computing system;

measure the change in the performance metric based on the application of the upgrade to the target computing system; and

update the machine learning system based on the measured change in the performance metric.

20. The computing system of claim 19 wherein the machine learning system is configured to identify the other computing system based on

identification of the customization to the computing process in the other computing system.