Minimizing impact of migrating virtual services

Patent metadata

Field	Value
Publication number	US-11567795-B2
Application number	US-202117217803-A
Country	US
Kind code	B2
Filing date	Mar 30, 2021
Priority date	Oct 15, 2018
Publication date	Jan 31, 2023
Grant date	Jan 31, 2023

Abstract

The present disclosure relates to systems, methods, and computer readable media that utilize a low-impact live-migration system to reduce unfavorable impacts caused as a result of live-migrating computing containers between physical server devices of a cloud computing system. For example, systems disclosed herein evaluates characteristics of computing containers on server devices to determine a predicted unfavorable impact of live-migrating the computing containers between the server devices. Based on the predicted impact, the systems disclosed herein can selectively identify which computing containers to live-migrate as well as carry out live-migration of the select computing containers in such a way the significantly reduces unfavorable impacts to a customer or client device associated with the computing containers.

First claim

What is claimed is:

1. A method for live-migrating virtual services between server nodes, comprising:
determining impact scores for a plurality of virtual machines on a cloud computing system, wherein determining the impact scores includes, for each virtual machine from the plurality of virtual machines, determining an impact score indicating a predicted impact of live-migrating a virtual machine based on one or more behavior characteristics of the virtual machine, the one or more behavior characteristics of the virtual machine including a sensitivity of the virtual machine to an estimated brownout time of live-migrating the virtual machine, the estimated brownout time indicating an estimated duration of time in which the virtual machine will experience a limited connection while live-migrating the virtual machine to a destination server device, wherein the estimated brownout time is determined to fall within a first brownout time range of a plurality of estimated brownout time ranges, and wherein the impact score is based on the sensitivity of the virtual machine to the first brownout time range;
identifying candidate virtual machines for live-migration based on a policy of the cloud computing system associated with increasing availability of computing resources on server devices of the cloud computing system, wherein identifying the candidate virtual machines includes identifying a subset of virtual machines from the plurality of virtual machines based on associated impact scores for the subset of virtual machines being less than impact scores for other virtual machines from the plurality of virtual machines; and
initiating live-migration of the candidate virtual machines to the destination server device.

2. The method of claim 1 , wherein the one or more behavior characteristics includes characteristics of an application running on the virtual machine and an associated level of tolerance of the application to a brownout event of the estimated brownout time.

3. The method of claim 1 , wherein determining the impact score for each virtual machine from the plurality of virtual machines includes:
applying a brownout prediction engine to the plurality of virtual machines, wherein the brownout prediction engine is a machine learning model trained to predict an estimated brownout time for live-migrating each virtual machine from the plurality of virtual machines, wherein the estimated brownout time includes the estimated duration of time that a virtual machine will experience the limited connection while still maintaining a level of client access to the virtual machine.

4. The method of claim 1 , wherein determining the impact score for each virtual machine from the plurality of virtual machines includes:
identifying lifetimes of the plurality of virtual machines, the lifetimes including indications of when the plurality of virtual machines are expected to expire; and
wherein determining the impact score for the virtual machine is further based on an associated lifetime of the virtual machine.

5. The method of claim 4 , wherein identifying lifetimes of the plurality of virtual machines includes identifying times at which one or more of the virtual machines are scheduled to expire.

6. The method of claim 4 , wherein determining the impact score for the virtual machine includes:
comparing the associated lifetime to a time when the virtual machine would be scheduled for live-migration; and
determining the impact score for the virtual machine based on comparing the associated lifetime to the time when the virtual machine would be scheduled for live-migration.

7. The method of claim 6 , wherein determining the impact score includes:
if the associated lifetime is predicted to expire earlier than the time when the virtual machine would be scheduled for live-migration, associating the virtual machine with a first impact score; and
if the associated lifetime is predicted to expire later than the time when the virtual machine would be schedule for live-migration, associating the virtual machine with a second impact score.

8. The method of claim 7 , wherein the first impact score is greater than the second impact score based on the associated lifetime being earlier than the time when the virtual machine would be scheduled for live-migration.

9. The method of claim 1 , wherein determining the impact score for each virtual machine includes:
identifying a time at which a virtual machine would be scheduled for live-migration; and
determining, based on the daily or weekly utilization pattern, the impact score based on a predicted utilization for the virtual machine at the time when the virtual machine would be scheduled for live-migration.

10. The method of claim 1 , wherein the one or more behavior characteristics further includes:
characteristics of an application running on the virtual machine;
a lifetime of the virtual machine, the lifetime including an indication of when the virtual machine is expected to expire; and
a daily or weekly utilization pattern for the virtual machine.

11. A system for live-migrating virtual services between server nodes, comprising:
one or more processors;
memory in electronic communication with the one or more processors; and
instructions stored in the memory, the instructions being executable by the one or more processors to:
determine impact scores for a plurality of virtual machines on a cloud computing system, wherein determining the impact scores includes, for each virtual machine from the plurality of virtual machines, determining an impact score indicating a predicted impact of live-migrating a virtual machine based on one or more behavior characteristics of the virtual machine, the one or more behavior characteristics of the virtual machine including a sensitivity of the virtual machine to an estimated brownout time of live-migrating the virtual machine, the estimated brownout time indicating an estimated duration of time in which the virtual machine will experience a limited connection while live-migrating the virtual machine to a destination server device;
identify candidate virtual machines for live-migration based on a policy of the cloud computing system associated with increasing availability of computing resources on server devices of the cloud computing system, wherein identifying the candidate virtual machines includes identifying a subset of virtual machines from the plurality of virtual machines based on associated impact scores for the subset of virtual machines being less than impact scores for other virtual machines from the plurality of virtual machines, wherein the sensitivity of the virtual machine to the estimated brownout time is based on a type of application hosted by the virtual machine and an associated level of tolerance of the type of application to a brownout event of the estimated brownout time; and
initiate live-migration of the candidate virtual machines to the destination server device.

12. The system of claim 11 , wherein the one or more behavior characteristics further includes:
characteristics of an application running on the virtual machine;
a lifetime of the virtual machine, the lifetime including an indication of when the virtual machine is expected to expire; and
a daily or weekly utilization pattern for the virtual machine.

13. The system of claim 11 , wherein determining the impact score for each virtual machine from the plurality of virtual machines includes:
applying a brownout prediction engine to the plurality of virtual machines, wherein the brownout prediction engine is a machine learning model trained to predict an estimated brownout time for live-migrating each vi

Assignees

• Microsoft Technology Licensing Llc

Inventors

• Cortez Eli
• Mani Ajay
• Fontoura Marcus Felipe
• Sheth Nisarg Tarakkumar
• Moscibroda Thomas
• Constantin Ana-Maria

Classifications

• G06F2009/45595

  Network integration; Enabling network access in virtual machine instances · CPC title

• H04L67/10

  in which an application is distributed across nodes in the network (software deployment G06F8/60; multiprogramming arrangements G06F9/46) · CPC title

• G06F9/45558 Primary

  Hypervisor-specific management and integration aspects · CPC title

• G06F2009/4557

  Distribution of virtual machine instances; Migration and load balancing · CPC title

• G06N5/02

  Knowledge representation; Symbolic representation · CPC title