Virtual machine power consumption measurement and management

Patent metadata

Field	Value
Publication number	US-9575539-B2
Application number	US-201414452451-A
Country	US
Kind code	B2
Filing date	Aug 5, 2014
Priority date	Feb 26, 2010
Publication date	Feb 21, 2017
Grant date	Feb 21, 2017

Abstract

Embodiments of the virtual machine power metering system and method measure the power consumption of individual virtual machines. Power meter measurements for a physical host server are converted into individual virtual machine power meters that measure the power consumption of each individual virtual machine residing on the host server. The virtual machine power consumption is computed by generating a power model using the total power consumption of the host server and resource utilization for a virtual machine. Optimal power model coefficients are computed using the power model. The energy used by the virtual machine is computed using one of two embodiments. Embodiments of the system and method also can be used to obtain the power consumption for a specific activity (such as a service, request, or search query). In addition, the virtual machine power metering can be used for virtual machine power capping to allow power oversubscription in virtualized environments.

First claim

What is claimed is:

1. A method comprising:
monitoring power consumptions of a plurality of virtual machines on a host server;
measuring peak power draws by the plurality of virtual machines on the host server;
determining whether an individual virtual machine has exceeded a corresponding virtual machine power budget of the individual virtual machine; and
in an instance when the individual virtual machine has exceeded the corresponding virtual machine power budget:
determining whether a host server power budget of the host server has been exceeded; and
responsive to a determination that the host server power budget has been exceeded, reducing a power consumption of the individual virtual machine by reducing an amount of processor time allocated to the individual virtual machine.

2. The method of claim 1 , the reducing the amount of processor time allocated to the individual virtual machine being performed without reducing another amount of processor time allocated to another virtual machine that shares a processor with the individual virtual machine.

3. The method of claim 1 , further comprising:
generating a power model that relates a total resource utilization for each of the plurality of virtual machines to a total power consumption of the host server;
using the power model to obtain optimal power model coefficients; and
computing the power consumptions of the plurality of virtual machines using the power model and the optimal power model coefficients.

4. The method of claim 1 , further comprising reducing power consumption of at least one other virtual machine that is exceeding a corresponding other virtual machine power budget so as to not exceed the host server power budget of the host server.

5. The method of claim 1 , further comprising:
in another instance when the individual virtual machine has exceeded the corresponding virtual machine power budget:
determining whether the host server power budget has been exceeded; and
responsive to another determination that the host server power budget has been exceeded, reducing the power consumption of the host server by reducing an amount of processor time allocated to another individual virtual machine that shares a processor with the individual virtual machine.

6. The method of claim 5 , the another individual virtual machine having a relatively lower priority than the individual virtual machine.

7. The method of claim 1 , further comprising migrating one or more of the plurality of virtual machines to another host server so as to not exceed the host server power budget.

8. A computing system comprising:
one or more hardware resources including one or more processing units; and
one or more computer-readable media storing computer readable instructions that, when executed by the one or more processing units, cause the one or more processing units to:
measure hardware resource utilization of hardware resources by multiple virtual machines that execute on the computing system;
based at least on the measured hardware resource utilization, determine power consumption measurements of the multiple virtual machines that execute on the computing system;
determine that a first virtual machine of the multiple virtual machines has exceeded a first power budget of the first virtual machine; and
reduce power consumption of the computing system by reducing an amount of processor time allocated to the first virtual machine on an individual processing unit that is shared by the first virtual machine and a second virtual machine that has not exceeded a second power budget of the second virtual machine.

9. The computing system of claim 8 , wherein the computer readable instructions, when executed by the one or more processing units, cause the one or more processing units to:
based at least on the power consumption measurements, determine whether a physical power budget of the computing system will be exceeded;
throttle the first virtual machine when the physical power budget will be exceeded; and
allow the first virtual machine to execute without being throttled when the physical power budget will not be exceeded.

10. The computing system of claim 8 , wherein the computer readable instructions, when executed by the one or more processing units, cause the one or more processing units to:
detect a context switch on an individual processing unit from the first virtual machine to the second virtual machine; and
determine the power consumption measurements such that:
a first power consumption measurement for the first virtual machine reflects first power consumption by the first virtual machine while the first virtual machine was active on the individual processing unit prior to the context switch, and
a second power consumption measurement for the second virtual machine reflects second power consumption by the second virtual machine while the second virtual machine was active on the individual processing unit after the context switch.

11. The computing system of claim 8 , wherein the computer readable instructions, when executed by the one or more processing units, cause the one or more processing units to:
obtain a model relating the hardware resource utilization to power consumption; and
determine the power consumption measurements of the multiple virtual machines using the model.

12. The computing system of claim 11 , the model having a first coefficient for processing unit utilization and a second coefficient for storage utilization, the power consumption measurements reflecting power consumed by the one or more processing units and storage hardware.

13. The computing system of claim 12 , the model having a third coefficient for network utilization, the power consumption measurements reflecting power consumed by the one or more processing units, the storage hardware, and network hardware.

14. The computing system of claim 13 , provided as a host server having a power sensor configured to measure the power consumption of the host server.

15. A computing system comprising:
one or more hardware resources including one or more processing units; and
one or more computer-readable media storing computer readable instructions that, when executed by the one or more processing units, cause the one or more processing units to:
measure corresponding processor utilizations of multiple virtual machines when executing on the one or more processing units;
measure memory or storage utilizations of the multiple virtual machines;
determine power consumption measurements of the multiple virtual machines based at least on the processor utilizations and the memory or storage utilizations; and
reduce power consumption of the computing system by throttling less than all of the multiple virtual machines.

16. The computing system of claim 15 , further comprising a power sensor configured to measure overall power consumption of the computing system.

17. The computing system of claim 16 , wherein the computer readable instructions, when executed by the one or more processing units, cause the one or more processing units to:
based at least on the overall power consumption measured by the power sensor, determine whether an overall power budget of the computing system has been exceeded; and
throttle less than all of the virtual machines when the overall power budget of the computing system has been exceeded.

18. The computing system of claim 16 , wherein the computer readable instructions, when executed by the one or more processing units, cause the one or more processing units to:
based at least on th

Assignees

• Microsoft Corp
• Microsoft Technology Licensing Llc

Inventors

• Kansal Aman
• Liu Jie
• Burger Douglas C
• Bhattacharya Arka Aloke

Classifications

• G06F1/3203 Primary

  Power management, i.e. event-based initiation of a power-saving mode · CPC title

• G06F11/3423

  where the assessed time is active or idle time · CPC title

• G06F11/3466

  Performance evaluation by tracing or monitoring · CPC title

• G06F11/3447

  Performance evaluation by modeling · CPC title

• G06F2201/865

  Monitoring of software · CPC title