Multivariable control for power-latency management to support optimization of data centers or other systems

• US 10,054,997 B2
• Filed: 09/29/2015
• Issued: 08/21/2018
• Est. Priority Date: 04/23/2015
• Status: Active Grant

• Current Assignee: Honeywell International Inc.
• Original Assignee: Honeywell International Inc.
• Inventors: Stewart, Gregory E., Nazari, Sohail, Dooner, Robert Paul Cartier
• Primary Examiner(s): Patel, Nitin
• Application Number:
  US14/869,689
• Publication Number:
  US 20160313780A1
• Time in Patent Office: 1,057 Days
• Field of Search: None
• CPC Class Codes: G06F 1/3206 Monitoring of events, devices or parameters that trigger a change in power modality
• CPC Class Codes: G06F 1/329 by task scheduling
• CPC Class Codes: G06F 9/50 Allocation of resources, e.g. of the central processing unit [CPU]
• CPC Class Codes: G06F 9/5083 Techniques for rebalancing the load in a distributed system
• CPC Class Codes: G06F 9/5094 where the allocation takes into account power or heat criteria power management in computers in general G06F1/3203; thermal management in computers in general G06F1/206
• CPC Class Codes: Y02D 10/00 Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

A method includes identifying demand for computing resources provided by multiple computing devices and identifying operating states or modes for the computing devices based on the demand using a multivariable controller. The multivariable controller is configured to determine how to alter multiple manipulated variables in order to create changes to multiple controlled variables. The multiple manipulated variables include the operating states or modes of the computing devices, and the multiple controlled variables include a power consumption of the computing devices and a response time of the computing devices. Each of the computing devices could include one or more processing units, and each of the computing devices or processing units could be configured to selectively operate in one of the operating states or modes. The method could also include generating a profile identifying a number of computing devices or processing units to operate in each of the operating states or modes.

Claims

1. A method comprising:

identifying demand for computing resources provided by multiple computing devices, wherein the multiple computing devices correspond to multiple servers located at one or more location;

identifying operating states or modes for the computing devices based on the identified demand using a multivariable controller, the multivariable controller configured to determine how to alter multiple manipulated variables in order to create changes to multiple controlled variables;

generating one or more models using historical data;

correlating changes to the operating states or modes of the computing devices with changes in power consumption or with changes in response time, using the generated one or more models;

generating a profile identifying a number of computing devices or processing units to operate in each of the operating states or modes;

identifying one or more of the computing devices or processing units to undergo one or more state or mode changes based on the profile; and

generating one or more control signals to trigger the one or more state or mode changes in the one or more computing devices or processing units,

wherein the multiple manipulated variables include the operating states or modes of the computing devices; and

wherein the multiple controlled variables include a power consumption of the computing devices and a response time of the computing devices.

2. The method of claim 1, wherein the operating states or modes of the computing devices are identified in order to reduce or minimize the power consumption of the computing devices.

3. The method of claim 1, wherein:

each of the computing devices includes one or more processing units;

each of the computing devices or processing units is configured to selectively operate in one of the operating states or modes.

4. The method of claim 1, wherein the one or more control signals comprise at least one of:

a command to migrate a first specified number of processing units from a first performance state to another performance state;

a command to migrate a second specified number of processing units from another performance state to the first performance state;

a command to migrate a third specified number of processing units from the first performance state to a first standby/sleep state;

a command to migrate a fourth specified number of processing units from the first standby/sleep state to the first performance state; and

a command to migrate a fifth specified number of processing units from one standby/sleep state to another standby/sleep state.

5. The method of claim 1, wherein:

each of the computing devices includes one or more processing units; and

the controlled variables further comprise multiple aggregated performance indicators, each performance indicator identifying an aggregate performance characteristic for all of the computing devices or processing units in one of the operating states or modes.

6. The method of claim 1, wherein the multivariable controller handles the demand for the computing resources as a disturbance variable.

7. The method of claim 1, wherein the multivariable controller is configured to solve an optimization problem in order to reduce or minimize the power consumption of the computing devices while ensuring that the computing devices provide an amount of computing resources to satisfy the demand while providing a specified level of service to users of the computing resources.

8. An apparatus comprising:

a multivariable controller comprising at least one processing device configured to;

identify demand for computing resources provided by multiple computing devices, wherein the multiple computing devices correspond to multiple servers located at one or more location;

identify operating states or modes for the computing devices based on the identified demand;

generate one or more models using historical data;

correlate changes to the operating states or modes of the computing devices with changes in power consumption or with changes in response time, using the generated one or more models;

generate a profile identifying a number of computing devices or processing units to operate in each of the operating states or modes;

identify one or more of the computing devices or processing units to undergo one or more state or mode changes based on the profile; and

generate one or more control signals to trigger the one or more state or mode changes in the one or more computing devices or processing units,

wherein the at least one processing device is configured to identify the operating states or modes for the computing devices by determining how to alter multiple manipulated variables in order to create changes to multiple controlled variables;

wherein the multiple manipulated variables include the operating states or modes of the computing devices; and

wherein the multiple controlled variables include a power consumption of the computing devices and a response time of the computing devices.

9. The apparatus of claim 8, wherein the at least one processing device is configured to identify the operating states or modes of the computing devices in order to reduce or minimize the power consumption of the computing devices.

10. The apparatus of claim 8, wherein:

each of the computing devices includes one or more processing units;

each of the computing devices or processing units is configured to selectively operate in one of the operating states or modes.

11. The apparatus of claim 8, wherein:

each of the computing devices includes one or more processing units; and

the controlled variables further comprise multiple aggregated performance indicators, each performance indicator identifying an aggregate performance characteristic for all of the computing devices or processing units in one of the operating states or modes.

12. The apparatus of claim 8, wherein the multivariable controller is configured to solve an optimization problem in order to reduce or minimize the power consumption of the computing devices while ensuring that the computing devices provide an amount of computing resources to satisfy the demand while providing a specified level of service to users of the computing resources.

13. The apparatus of claim 12, wherein the multivariable controller is configured to:

identify the demand for the computing resources by predicting or receiving a prediction of estimated future workload on the computing devices; and

treat the estimated future workload as a disturbance variable when solving the optimization problem.

14. A non-transitory computer readable medium containing computer readable program code that, when executed, causes at least one processing device of a multivariable controller to:

identify demand for computing resources provided by multiple computing devices, wherein the multiple computing devices correspond to multiple servers located at one or more location;

identify operating states or modes for the computing devices based on the identified demand by determining how to alter multiple manipulated variables in order to create changes to multiple controlled variables;

generate one or more models using historical data;

correlate changes to the operating states or modes of the computing devices with changes in power consumption or with changes in response time, using the generated one or more models;

generate a profile identifying a number of computing devices or processing units to operate in each of the operating states or modes;

identify one or more of the computing devices or processing units to undergo one or more state or mode changes based on the profile; and

generate one or more control signals to trigger the one or more state or mode changes in the one or more computing devices or processing units,

wherein the multiple manipulated variables include the operating states or modes of the computing devices; and

wherein the multiple controlled variables include a power consumption of the computing devices and a response time of the computing devices.

15. The non-transitory computer readable medium of claim 14, wherein:

each of the computing devices includes one or more processing units;

each of the computing devices or processing units is configured to selectively operate in one of the operating states or modes.

16. The non-transitory computer readable medium of claim 14, wherein:

each of the computing devices includes one or more processing units;

the controlled variables further comprise multiple aggregated performance indicators, each performance indicator identifying an aggregate performance characteristic for all of the computing devices or processing units in one of the operating states or modes; and

the computer readable program code that, when executed, causes the at least one processing device to identify the operating states or modes for the computing devices comprises;

computer readable program code that, when executed, causes the at least one processing device to solve an optimization problem in order to reduce or minimize the power consumption of the computing devices while ensuring that the computing devices provide an amount of computing resources to satisfy the demand while providing a specified level of service to users of the computing resources.