Automated management of power distribution during a power crisis

Patent metadata

Field	Value
Publication number	US-2021397239-A1
Application number	US-202016906409-A
Country	US
Kind code	A1
Filing date	Jun 19, 2020
Priority date	Jun 19, 2020
Publication date	Dec 23, 2021
Grant date	—

Abstract

A method comprises analyzing performance data of a system using one or more machine learning techniques. The system comprises a plurality of hardware components. In the method, a priority list of the plurality of hardware components is generated based on the analysis, and power from one or more power sources is distributed to one or more of the plurality of hardware components based on the priority list.

First claim

1 . A method, comprising:
analyzing performance data of a system using one or more machine learning techniques, the system comprising a plurality of hardware components; generating a priority list of the plurality of hardware components based on the analysis; distributing power from one or more power sources to a first hardware component subset comprising one or more of the plurality of hardware components, wherein the distributing is based on the priority list; identifying a second hardware component subset comprising at least one of the plurality of hardware components that are without power following the distributing of the power from the one or more power sources to the first hardware component subset; and controlling power usage of the one or more of the plurality of hardware components of the first hardware component subset responsive to the identifying; wherein the steps of the method are executed by a processing device operatively coupled to a memory.

2 . The method of claim 1 , further comprising computing available power from the one or more of the power sources.

3 . The method of claim 1 , wherein the performance data comprises at least one of memory usage of the plurality of hardware components, one or more of applications, tasks and services running on the plurality of hardware components, system log data, system traffic data and system load data.

4 . The method of claim 1 , wherein the one or more machine learning techniques comprises a time series model, and wherein the analyzing comprises using the time series model to analyze at least one of: (i) one or more of capacity data and availability data of the plurality of hardware components; (ii) one or more of applications, tasks and services running on the plurality of hardware components; (iii) health data of the plurality of hardware components; and (iv) utilization data of the plurality of hardware components to determine real-time performance states of the plurality of hardware components.

5 . The method of claim 4 , wherein the time series model comprises an Autoregressive Integrated Moving Average (ARIMA) model.

6 . The method of claim 4 , wherein the utilization data comprises at least one of central processing unit (CPU) utilization, memory utilization, network utilization and storage utilization of the plurality of hardware components.

7 . The method of claim 1 , wherein the plurality of hardware components comprise a plurality of servers.

8 . The method of claim 1 ,
wherein the controlling of the power usage comprises reducing power consumption of the one or more of the plurality of hardware components of the first hardware component subset.

9 . The method of claim 8 , further comprising distributing power saved by the reduced power consumption to the at least one of the plurality of hardware components without power.

10 . The method of claim 8 , wherein the controlling of the power usage comprises at least one of reducing a fan speed of the one or more of the plurality of hardware components of the first hardware component subset, changing a thermal profile of the one or more of the plurality of hardware components of the first hardware component subset, and dynamically capping the power usage of the one or more of the plurality of hardware components of the first hardware component subset.

11 . The method of claim 1 , further comprising:
monitoring power consumption of the plurality of hardware components; identifying an increase in the power consumption by at least a first one of the plurality of hardware components; and powering off at least a second one of the plurality of hardware components having a lower priority on the priority list than a priority of at least the first one of the plurality of hardware components.

12 . The method of claim 1 , further comprising identifying a power crisis in the system, wherein the distributing of the power based on the priority list is performed in response to the identifying of the power crisis.

13 . An apparatus comprising:
a processing device operatively coupled to a memory and configured to: analyze performance data of a system using one or more machine learning techniques, the system comprising a plurality of hardware components; generate a priority list of the plurality of hardware components based on the analysis; distribute power from one or more power sources to a first hardware component subset comprising one or more of the plurality of hardware components, wherein the distributing is based on the priority list; identify a second hardware component subset comprising at least one of the plurality of hardware components that are without power following the distributing of the power from the one or more power sources to the first hardware component subset; and control power usage of the one or more of the plurality of hardware components of the first hardware component subset responsive to the identifying.

14 . The apparatus of claim 13 , wherein, in controlling the power usage, the processing device is configured to:
reduce power consumption of the one or more of the plurality of hardware components of the first hardware component subset.

15 . The apparatus of claim 14 , wherein the processing device is further configured to distribute power saved by the reduced power consumption to the at least one of the plurality of hardware components without power.

16 . The apparatus of claim 13 , wherein the processing device is further configured to:
monitor power consumption of the plurality of hardware components; identify an increase in the power consumption by at least a first one of the plurality of hardware components; and power off at least a second one of the plurality of hardware components having a lower priority on the priority list than a priority of at least the first one of the plurality of hardware components.

17 . An article of manufacture comprising a non-transitory processor-readable storage medium having stored therein program code of one or more software programs, wherein the program code when executed by at least one processing device causes said at least one processing device to perform the steps of:
analyzing performance data of a system using one or more machine learning techniques, the system comprising a plurality of hardware components; generating a priority list of the plurality of hardware components based on the analysis; distributing power from one or more power sources to a first hardware component subset comprising one or more of the plurality of hardware components, wherein the distributing is based on the priority list; identifying a second hardware component subset comprising at least one of the plurality of hardware components that are without power following the distributing of the power from the one or more power sources to the first hardware component subset; and controlling power usage of the one or more of the plurality of hardware components of the first hardware component subset responsive to the identifying.

18 . The article of manufacture of claim 17 , wherein, in controlling the power usage, the program code causes said at least one processing device to perform the step of:
reducing power consumption of the one or more of the plurality of hardware components of the first hardware component subset.

19 . The article of manufacture of claim 18 , wherein the program code further causes said at least one processing device to perform the step of distributing power saved by the reduced power consumption to the at least one of the pl

Assignees

• Dell Products Lp

Inventors

• Sethi Parminder Singh
• Kumar Suren

Classifications

• Y02D10/00

  Energy efficient computing, e.g. low power processors, power management or thermal management · CPC title

• G06N20/00

  Machine learning · CPC title

• G06F1/3287

  by switching off individual functional units in the computer system · CPC title

• G06F1/206

  comprising thermal management · CPC title

• G06F1/30 Primary

  Means for acting in the event of power-supply failure or interruption, e.g. power-supply fluctuations (for resetting only G06F1/24) · CPC title