Operational management in cloud computing systems

• US 10,250,456 B2
• Filed: 05/17/2016
• Issued: 04/02/2019
• Est. Priority Date: 05/17/2016
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Barrett, Dan
• Primary Examiner(s): Winder, Patrice L
• Assistant Examiner(s): Aung, Aye M
• Application Number:
  US15/157,204
• Publication Number:
  US 20170339027A1
• Time in Patent Office: 1,050 Days
• Field of Search: 709224
• CPC Class Codes: G06F 9/5072 Grid computing
• CPC Class Codes: H04L 41/5009 Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF]
• CPC Class Codes: H04L 43/0823 Errors, e.g. transmission errors
• CPC Class Codes: H04L 43/14 using software, i.e. software packages network security related monitoring H04L63/1408
• CPC Class Codes: H04L 43/16 Threshold monitoring
• CPC Class Codes: H04L 49/35 Switches specially adapted for specific applications
• CPC Class Codes: H04L 67/10 in which an application is distributed across nodes in the network software deployment G06F8/60; multiprogramming arrangements G06F9/46

Abstract

Techniques for managing operation in cloud computing systems are disclosed herein. In one embodiment, a method can include receiving data representing a guaranteed value of a performance metric of a cloud service and an error budget and deriving a switching threshold based on a combination of the value of the performance metric and the error budget. The method also includes determining a current value of the performance metric of the cloud service and causing the cloud computing system to selectively switch between operational modes for providing the cloud service based on a comparison between the determined current value of the performance metric and the switching threshold.

Claims

1. A method of managing operations in a cloud computing system having one or more software applications deployed on one or more servers of the cloud computing system and configured to provide a cloud service, the method comprising:

receiving, at a control server, data representing a guaranteed value of a performance metric of the cloud service according to a service level agreement and data representing an error budget associated with the performance metric;

deriving, at the control server, a switching threshold based on a combination of the received data representing the value of the performance metric and the error budget;

determining, at the control server, a current value of the performance metric of the cloud service provided by the cloud computing system;

determining whether the current value of the performance metric exceeds the derived switching threshold; and

in response to determining that the current value of the performance metric exceeds the switching threshold, with the control server, deploying one or more remedial packages to the one or more servers at which the software applications are deployed while disallowing submission, configuration, or deployment of any new build of the one or more software applications to any of the one or more servers in the cloud computing system, the remedial packages being configured to improve performance of the already deployed one or more software applications, thereby delivering the cloud service at or above the guaranteed value of the performance metric according to the service level agreement.

2. The method of claim 1 wherein deriving the switching threshold includes subtracting both the guaranteed value of the performance metric and the error budget from another value of the performance metric corresponding to an error-free operation of the cloud service.

3. The method of claim 2 wherein:

determining the current value of the performance metric includes calculating an error rate of the performance metric for the cloud service; and

determining whether the current value of the performance metric exceeds the derived switching threshold includes determining whether the calculated error rate exceeds the switching threshold.

4. The method of claim 2 wherein:

determining the current value of the performance metric includes calculating an error rate of the performance metric for the cloud service;

determining whether the current value of the performance metric exceeds the derived switching threshold includes determining whether the calculated error rate exceeds the switching threshold; and

the method further includes allowing submission, configuration, or deployment of a new build of the one or more software applications to the one or more servers in the cloud computing system in response to determining that the calculated error rate does not exceed the switching threshold.

5. The method of claim 2 wherein the method further includes:

maintaining an operational mode of the provided cloud service in response to determining that the current value of the performance metric is equal or greater than the switching threshold.

6. The method of claim 1 wherein deriving the switching threshold includes setting the switching threshold to a sum of the guaranteed value of the performance metric and the error budget.

7. The method of claim 1 wherein:

deriving the switching threshold includes setting the switching threshold to a sum of the guaranteed value of the performance metric and the error budget; and

determining whether the current value of the performance metric exceeds the switching threshold includes determining whether the current value of the performance metric exceeds the sum of the guaranteed value of the performance metric and the error budget.

8. The method of claim 1 wherein the method further includes:

preventing allocation of additional bandwidth to the cloud service in response to determining that the current value of the performance metric exceeds the switching threshold.

9. The method of claim 1 wherein the method further includes allowing a new build of the software application to be deployed to the one or more servers in response to determining that the current value of the performance metric does not exceed the switching threshold.

10. A computing device, comprising:

a processor; and

a memory containing instructions executable by the processor to cause the computing device to;

in a cloud computing system, monitor a status of a cloud service provided by the cloud computing system having one or more software applications deployed on one or more servers in the cloud computing system to provide the cloud service;

based on the monitored status of the cloud service, determine a current value of a performance metric of the cloud service;

perform a comparison between the determined current value of the performance metric and a switch threshold generated based on a combination of a guaranteed value of the performance metric of the cloud service according to a service level agreement and an adjustable error budget; and

in response to detecting that the current value of the performance metric exceeds the switching threshold based on the performed comparison, deploy one or more remedial packages to the one or more servers at which the software applications are deployed while disallowing submission, configuration, or deployment of new build of the software applications to any of the one or more servers in the cloud computing system, the remedial packages being configured to improve performance of the already deployed software applications.

11. The computing device of claim 10 wherein the switching threshold is generated by subtracting both the guaranteed value of the performance metric and the error budget from another value of the performance metric corresponding to an error-free operation of the cloud service.

12. The computing device of claim 11 wherein:

to determine the current value of the performance metric includes to calculate an error rate of the performance metric for the cloud service;

to perform the comparison includes to determine whether the calculated error rate exceeds the switching threshold.

13. The computing device of claim 11 wherein:

to determine the current value of the performance metric includes to calculate an error rate of the performance metric for the cloud service;

to perform the comparison includes to determine whether the calculated error rate exceeds the switching threshold; and

the memory contains additional instructions executable by the processor to cause the computing device to, in response to determining that the calculated error rate does not exceed the switching threshold, allow submission, configuration, or deployment of one or more new builds of the software application to the one or more servers.

14. The computing device of claim 10 wherein the switching threshold is set to a sum of the guaranteed value of the performance metric and the error budget.

15. The computing device of claim 14 wherein:

to perform the comparison includes to determine whether the current value of the performance metric is lower than the switching threshold; and

in response to determining that the current value of the performance metric exceeds the switching threshold, maintaining the cloud computing system in an operational mode under which submission, configuration, or deployment of one or more new builds of the software application is allowed.

16. The computing device of claim 14 wherein:

to perform the comparison includes to determine whether the current value of the performance metric is lower than the switching threshold; and

in response to determining that the current value of the performance metric is lower than the switching threshold, to deploy the one or more remedial packages to the one or more servers at which the software applications are deployed while disallowing submission, configuration, or deployment of new build of the software applications to any servers in the cloud computing system.

17. A method of managing operations in a cloud computing system having one or more software applications deployed on one or more servers of the cloud computing system that are managed by a control server, execution of the one or more software applications providing a cloud service, the method comprising:

determining, at the control server, a current value of the performance metric of the cloud service provided by the cloud computing system;

determining whether the current value of the performance metric exceeds a switching threshold based on a combination of a guaranteed value of the performance metric of the cloud service according to a service level agreement and an error budget associated with the performance metric; and

in response to determining that the current value of the performance metric exceeds the switching threshold, with the control server, deploying one or more remedial packages to the one or more servers at which the software applications are deployed while disallowing submission, configuration, or deployment of any new build of the software applications to any of the one or more servers in the cloud computing system, the remedial packages being configured to improve performance of the already deployed software applications, thereby delivering the cloud service at or above the guaranteed value of the performance metric according to the service level agreement.

18. The method of claim 17 wherein deriving the switching threshold includes subtracting both the guaranteed value of the performance metric and the error budget from another value of the performance metric corresponding to an error-free operation of the cloud service.

19. The method of claim 18 wherein:

determining the current value of the performance metric includes calculating an error rate of the performance metric for the cloud service; and

determining whether the current value of the performance metric exceeds the derived switching threshold includes determining whether the calculated error rate exceeds the switching threshold.

20. The method of claim 18 wherein:

determining the current value of the performance metric includes calculating an error rate of the performance metric for the cloud service;

determining whether the current value of the performance metric exceeds the derived switching threshold includes determining whether the calculated error rate exceeds the switching threshold; and

the method further includes allowing submission, configuration, or deployment of a new build of the software applications to the one or more servers in the cloud computing system in response to determining that the calculated error rate does not exceed the switching threshold.