CN114402272A - Power management method and apparatus - Google Patents
Power management method and apparatus Download PDFInfo
- Publication number
- CN114402272A CN114402272A CN201980099694.XA CN201980099694A CN114402272A CN 114402272 A CN114402272 A CN 114402272A CN 201980099694 A CN201980099694 A CN 201980099694A CN 114402272 A CN114402272 A CN 114402272A
- Authority
- CN
- China
- Prior art keywords
- power capping
- power
- mode
- capping
- computing system
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000007726 management method Methods 0.000 title description 44
- 238000000034 method Methods 0.000 claims abstract description 45
- 230000004044 response Effects 0.000 claims abstract description 33
- 238000012544 monitoring process Methods 0.000 claims description 25
- 230000008569 process Effects 0.000 description 21
- 238000010586 diagram Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 7
- 238000012545 processing Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 5
- 230000007246 mechanism Effects 0.000 description 4
- 238000004088 simulation Methods 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 3
- 230000003111 delayed effect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 230000007723 transport mechanism Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
- G06F1/26—Power supply means, e.g. regulation thereof
- G06F1/32—Means for saving power
- G06F1/3203—Power management, i.e. event-based initiation of a power-saving mode
- G06F1/3234—Power saving characterised by the action undertaken
- G06F1/3243—Power saving in microcontroller unit
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
- G06F1/26—Power supply means, e.g. regulation thereof
- G06F1/32—Means for saving power
- G06F1/3203—Power management, i.e. event-based initiation of a power-saving mode
- G06F1/3234—Power saving characterised by the action undertaken
- G06F1/3287—Power saving characterised by the action undertaken by switching off individual functional units in the computer system
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
- G06F1/26—Power supply means, e.g. regulation thereof
- G06F1/32—Means for saving power
- G06F1/3203—Power management, i.e. event-based initiation of a power-saving mode
- G06F1/3234—Power saving characterised by the action undertaken
- G06F1/329—Power saving characterised by the action undertaken by task scheduling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/48—Program initiating; Program switching, e.g. by interrupt
- G06F9/4806—Task transfer initiation or dispatching
- G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
- G06F9/4881—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
- G06F9/4893—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues taking into account power or heat criteria
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Computer Hardware Design (AREA)
- Computing Systems (AREA)
- Power Sources (AREA)
Abstract
Methods and apparatus for improved power management are provided. A power management component receives a power capping command. The power capping command includes a power capping target of the computing system. The power management component determines whether a power consumption of the computing system satisfies the power capping target when performing the power capping in one of a first power capping mode and a second power capping mode. The power management component performs the power capping in the first power capping mode and the second power capping mode in response to determining that the power consumption of the computing system does not meet the power capping target when performing the power capping in one of the first power capping mode and the second power capping mode.
Description
Technical Field
The present disclosure relates to the field of power management, and more particularly, to methods and apparatus for coordinating different power capping modes.
Background
Power capping is a technique widely used in modern Data Centers (DC) to increase on-rack computation density and avoid blackouts. Hardware (e.g., chip, server) vendors such as Original Design Manufacturers (ODMs) and Original Equipment Manufacturers (OEMs) as well as software (e.g., Operating System (OS), virtualization hypervisor) vendors provide basic power capping capabilities.
However, the vendor does not place the power capping function blocks together into a solution that is able to coordinate different power capping modes to achieve the power capping target.
Disclosure of Invention
This summary is not intended to identify essential features of the claimed subject matter, nor is it intended for use in limiting the scope of the claimed subject matter.
Example implementations of power management are described below. In an implementation, a power management component receives a power capping command. The power capping command includes a power capping target for the computing system. The power management component determines whether the power consumption of the computing system meets a power capping target when performing power capping in one of the first power capping mode and the second power capping mode. The power management component performs power capping in the first power capping mode and the second power capping mode in response to determining that the power consumption of the computing system does not meet the power capping target when performing power capping in one of the first power capping mode and the second power capping mode. The first power capping mode is an out-of-band (OOB) power capping mode and the second power capping mode is an in-band (IB) power capping mode.
Thus, the OOB power capping mode and the IB power capping mode are coordinated and combined to achieve the power capping target of the computing system. Thus, power capping is more reliable and usable.
Drawings
The detailed description is set forth with reference to the accompanying drawings. In the figures, the left-most digit(s) of a reference number identifies the figure in which the reference number first appears. The use of the same reference symbols in different drawings indicates similar or identical items or features.
FIG. 1 illustrates an example block diagram of a computing system.
Fig. 2 illustrates an example flow diagram of a process for coordinating different power capping modes.
Fig. 3A, 3B, and 3C illustrate in detail example flow diagrams of processes for coordinating different power capping modes.
FIG. 4 illustrates an example device for implementing the above-described processes and methods.
Detailed Description
The terms as used herein are expressed as follows.
Out-of-band (OOB) refers to a method of accessing a computing system without using the Operating System (OS) of the computing system.
In-band (IB) refers to a method of accessing a computing system via its OS.
Management Engine (ME) refers to logic typically residing in a CPU external bridge chip on an intel platform.
A Node Manager (NM) refers to a piece of firmware or state machine in the ME that is responsible for power management of the computing system.
A telemetry data hub refers to a piece of firmware or state machine in the ME that is responsible for collecting performance telemetry data from the processor.
The rest of the platform (RoP) refers to other computing components than the main parts like CPU, GPU, Platform Controller Hub (PCH), fan, Power Supply Unit (PSU).
Power capping refers to a method of limiting the power consumption of a computing system to not exceed a power limit/ceiling.
Board Management Controller (BMC) refers to a small embedded coprocessor that operates as an Intelligent Platform Management Interface (IPMI) bus master to perform system level power management, monitoring, logging, alarms, etc.
Fig. 1 illustrates an example block diagram of a computing system 100. The computing system may be implemented in a large-scale distributed computing environment.
Referring to fig. 1, computing system 100 includes a RoP 102, a storage device 104, a network 106, a fan 108, a Graphics Processing Unit (GPU)/Field Programmable Gate Array (FPGA)110, a double data rate synchronous dynamic random access memory (DDR SDRAM))112, a Central Processing Unit (CPU)114, a chipset 116, a BMC 118, an ME 120, a power capping pcode/ucode 122, a power management component 124, and an authentication 126. Computing system 100 is configured to get/set commands from a power cap master (not shown). In an implementation, the power capping master may be a controller, console, or the like in communication with the computing system 100.
The power management component 124 includes 3 parts, part a128, part B130, and part C132. Part a128 is an OOB power capping subagent. Part B130 is a coordinator. Part C132 is an IB power capping subagent. In implementations, the power management component 124 can be implemented in software, hardware, firmware, or any combination thereof.
The power management component 124 is configured to perform different power capping modes, e.g., IB power capping mode and OOB power capping mode. In IB power capping mode, power management component 124 is configured to run (as a service) or ready to run on the OS of computing system 100 in response to a management command. In the OOB power capping mode, the power management component 124 is configured to access a node manager (not shown) in the ME 120 using the management port. In that case, power management component 124 is not running on the OS of computing system 100.
The power management component 124 is further configured to communicate with the BMC 118 via a network interface to receive power management commands and send responses. The power management component 124 is further configured to access Mode Specific Registers (MSRs) to read/write power capping pcode/ucode 122. The power management component 124 is further configured to communicate with an authentication 128.
The OOB power capping subagent 128 is configured to perform power capping in the OOB power capping mode.
IB power capping subagent 132 is configured to perform power capping in IB power capping mode.
The coordinator 130 is configured to coordinate/dispatch the power capping tasks to the OOB power capping subagent 128 and/or the IB power capping subagent 132. The coordinator 130 is further configured to calculate the OOB success rate R over a period of time (e.g., a recent period) upon receiving the power capping commandOOB% and IB success rate ROOB% and assigns the power capping task to the OOB power capping subagent 128 or the IB power capping subagent 132 or both to reach the power capping target.
In an implementation, the OOB success rate may be calculated as follows. Within period T, when OOB power capping mode is executed on the computing system, the total number of attempts for power capping is N, and the number of attempts by the computing system to meet the power capping target for power consumption is N. OOB success rate ROOB％＝n/N*100％。
In an implementation, the IB success rate may be calculated as follows. Within time period T, when the IB power capping mode is executed on the computing system, the total number of attempts for power capping is N, and the number of attempts by the computing system to meet the power capping target for power consumption is N. IB success rate RIB％＝n/N*100％。
In an implementation, IB power capping mode or OOB power capping mode may be delayed or even rejected when traffic and/or load is high. For example, in a heavily used computer, the CPU utilization is 100%, the memory bandwidth utilization is 80% +, the store queue is 100% + full, and the network bandwidth utilization is 80% +. In OOB power capping mode, the power management component needs to pass IPMI protocol over the power management bus (PMBus) (not shown) with the help of the BMC as an embedded on-board co-processor. When there are a large number of hardware errors, the BMC needs to handle high priority activities such as logging and alerts, but can ignore, defer or reject medium priority activities such as power capping commands. A similar situation may occur in IB power capping mode. When utilization at the computing system is high, there is a chance that IB access is severely delayed or even suspended.
Different dimensions of cloud services, such as availability, latency, reliability, throughput, etc., can be adjusted based on Service Level Agreements (SLAs) between service providers and customers. Typically, cloud services need to provide up to 99.99% or even better availability. Software and applications may use distributed and/or redundant designs to achieve this level of availability on unreliable infrastructure. The practice of ultra-scale cloud services provides for increased density of server deployments, which significantly increases the risk of power outages. The service provider may set a power buffer large enough to prevent the risk of power outage. Additionally, power capping is performed to control power consumption to reduce the risk of power outages. During power capping, the desired availability may be low, e.g., 99.9%. In an implementation, the availability in OOB power capping mode may be around 95%, which may vary under different conditions, but well below the desired availability of 99.9%.
In implementations, IB power capping mode and OOB power capping mode may be coordinated and combined to achieve power capping targets for the computing system. Further, availability of cloud services under power cap may be improved.
Fig. 2 illustrates an example flow diagram of a process 200 for coordinating different power capping modes.
In block 202, the power management component receives a power capping command. The power capping command includes a power capping target for the computing system. The power capping target may be dynamically set and/or adjusted based on actual needs.
In block 204, the power management component determines whether the power consumption of the computing system meets a power capping target when performing power capping in one of the first power capping mode and the second power capping mode.
In an implementation, a power management component determines whether a first power contribution of a first power capping mode is greater than a first nominal success rate. The power management component performs power capping in the first power capping mode in response to determining that the first power contribution of the first power capping mode is greater than the first nominal success rate. The first contribution power may be calculated as follows. During time period T, when the first power capping mode is executed on the computing system, the total number of attempts for power capping is N, and the number of attempts by the computing system to power consumption to meet the power capping target is N. The first power contribution is N/N x 100%. The first nominal success rate may be determined by, but is not limited to, specifications, historical data, statistical data, test data, simulation data, empirical data, and the like. The power management component waits until the first validity delay expires. The first effective delay is a period of time that the first power capping mode takes effect after the first power capping mode is implemented/executed. The first effective delay may be dynamically set and/or adjusted based on actual needs. The power management component determines whether the power consumption of the computing system meets a power capping target when performing power capping in the first power capping mode.
In an implementation, the power management component determines whether a second power of the second power capping mode is greater than a second nominal success rate. The second power contribution can be calculated as follows. During period T, when the second power capping mode is executed on the computing system, the total number of attempts for power capping is N, and the number of attempts by the computing system to meet the power capping target for power consumption is N. The second power generation is N/N x 100%. The second nominal success rate may be determined by, but is not limited to, specifications, historical data, statistical data, test data, simulation data, empirical data, and the like. The power management component performs power capping in the second power capping mode in response to determining that a second power contribution of the second power capping mode is greater than a second nominal success rate. The power management component waits until the second validity delay expires. The second effective delay is a period of time that the second power capping mode takes effect after the second power capping mode is implemented/executed. The second effective delay may be dynamically set and/or adjusted based on actual needs. The power management component determines whether the power consumption of the computing system meets a power capping target when performing power capping in the second power capping mode.
In an embodiment, the first power capping mode is an OOB power capping mode and the second power capping mode is an IB power capping mode.
In block 206, the power management component performs power capping in the first power capping mode and the second power capping mode in response to determining that the power consumption of the computing system does not meet the power capping target when performing power capping in one of the first power capping mode and the second power capping mode. For example, power capping may be performed in the first power capping mode and the second power capping mode simultaneously.
In an implementation, the power management component waits until both the first effective delay and the second effective delay expire, and determines whether the power consumption of the computing system meets a power capping target.
In an implementation, a power management component monitors whether a conflict occurs and stops performing power capping in one of a first power capping mode and a second power capping mode in response to the monitoring that the conflict occurs.
In an implementation, the power management component checks if a conflict occurs in the pcode/ucode setting.
With the example process 200 described above, the first power capping mode and the second power capping mode are coordinated and combined to achieve a power capping target for the computing system. Thus, power management during power capping is improved.
Fig. 3A, 3B, and 3C illustrate in detail an example flow diagram of a process 300 for coordinating different power capping modes.
At block 302, the coordinator receives a power capping command. The power capping command includes a power capping target for the computing system. The power capping target may be dynamically set and/or adjusted based on actual needs.
At block 304, the coordinator determines the OOB success rate ROOB% greater than nominal OOB success rate ROOB% of the total weight of the composition. OOB success rate ROOB% algorithm is as described above with reference to fig. 1. The nominal OOB success rate R may be determined by, but is not limited to, specifications, historical data, statistical data, test data, simulation data, empirical data, and the likeOOB*％。
If the coordinator determines the OOB success rate R at block 304OOB% greater than nominal OOB success rate ROOBThen, in block 306, the coordinator instructs the OOB power capping subagent to perform power capping in the OOB power capping mode with the node manager.
If the coordinator determines the OOB success rate R at block 304OOB% success rate R not greater than nominal OOBOOBAnd the process proceeds to block 316.
In block 308, the coordinator waits until the OOB effective delay TOOBUntil expiration. OOB effective delay TOOBIs a period of time during which the OOB power capping mode is in effect after the OOB power capping mode is implemented/executed. The OOB effective delay T may be dynamically set and/or adjusted based on actual needsOOB。
At block 310, the coordinator checks the power consumption P of the computing system1。
At block 312, the coordinator determines the power consumption P of the computing system1Whether less than or equal to the power capping target.
If the coordinator determines that computing system P is present at block 3121Is less than or equal to the power capping target, the coordinator updates the OOB success rate R at block 314OOB％。
If the coordinator determines the power consumption P of the computing system at block 3121Not less than or equal to the power capping target, the coordinator prepares an initial value for the IB power capping mode at block 316. In an implementation, the initial value includes an IB success rate RIB% and nominal IB success rate RIB*％。
At block 318, the process 300 is complete.
As shown in FIG. 3B, the process 300 continues from block 316, and at block 320, the coordinator determines IB success rate RIB% success rate R over nominal IBIB% of the total weight of the composition. The algorithm for IB success rate RIB% is as described above with reference to fig. 1. The nominal IB success rate R may be determined by, but is not limited to, specifications, historical data, statistical data, test data, simulation data, empirical data, and the likeIB*％。
If the coordinator determines at block 320 that the IB success rate R is highIB% greater than nominal IB success rate RIBThen the coordinator instructs the IP power capping subagent to perform power capping in IB power capping mode at block 322.
At block 324, the coordinator waits until the IB effective delay TIBUntil expiration. IB effective delay TIBIs the period of time that IB power capping mode takes effect after it is implemented/executed. The IB effective delay T may be dynamically set and/or adjusted based on actual needsIB。
If the coordinator determines at block 320 that the IB success rate R is highIB% success rate R not greater than nominal IBIBThen, at block 326, the coordinator instructs the OOB power capping subagent to perform power capping in the OOB power capping mode with the node manager. At the same time, the coordinator also instructs the IP power capping subagent to perform power capping in IB power capping mode. In an implementation, the power management component performs power capping in both IB power capping mode and IB power capping mode simultaneously.
At block 328, the coordinator waits until the OOB effective delay TOOBAnd IB effective delay TIBAll expire.
After blocks 324 and 328, the process 300 proceeds to block 330.
At block 330, the coordinator determines the power consumption P of the computing system2Whether less than or equal to the power capping target.
If the coordinator determines the power consumption P of the computing system at block 3302Less than or equal to the power capping target, the coordinator updates the IB success rate R at block 332OOB％。
If the coordinator determines the work of the computing system at block 330Consumption P2Not less than or equal to the power capping target, then power capping is performed with an additional power control mechanism at block 334. In implementations, the additional power control mechanism may include an idle state control mechanism that controls an idle state of one or more processing components, an active state control mechanism that controls an active state of one or more processing components (e.g., frequency scaling during active periods), and so forth.
The coordinator waits at block 336 until the OOB + IB effective delay TBOTHUntil expiration. OOB + IB effective delay TBOTHIs the period of time during which power capping performed in both OOB power capping mode and IB power capping mode is in effect after both OOB power capping mode and IB power capping mode are implemented/performed. The OOB + IB effective delay may be dynamically set and/or adjusted based on actual needs.
At block 338, the coordinator determines the power consumption P of the computing system3Whether less than or equal to the power capping target.
If the coordinator determines the power consumption P of the computing system at block 3383Not less than or equal to the power capping target, the coordinator initiates a failover operational sequence flow at block 340. Additionally or alternatively, the coordinator may take an instance running on the computing system offline. Additionally or alternatively, instances running on a computing system may be migrated to other computing systems as soon as possible.
If the coordinator determines that computing system P is present at block 3383Is less than or equal to the power capping target, then the process 300 is complete at block 342.
After blocks 332, 338, and 340, the process 300 is complete at block 342.
Turning now to FIG. 3C and continuing from block 324, at block 344 the coordinator creates a monitoring process.
At block 346, the coordinator continues to check if there is a conflict in the pcode/ucode settings.
If the coordinator determines at block 346 that there is a conflict in the pcode/ucode settings, the coordinator stops performing power capping in one power capping mode at block 348. In an implementation, the coordinator stops performing power capping in the IB power capping mode when there is a conflict. However, the coordinator continues to perform power capping in the OOB power capping mode. The IB power capping mode is stopped first because the OOB power capping mode does not occupy the OS of the computing system. Alternatively, the coordinator may stop performing power capping in the OOB power capping mode while continuing to perform power capping in the OOB power capping mode.
If the coordinator determines at block 346 that there is no conflict in the pcode/ucode setting, the coordinator determines at block 350 an OOB effective delay TOOBAnd IB effective delay TIBWhether both are expired.
If the coordinator determines the OOB effective delay T at block 350OOBAnd IB effective delay TIBExpires, the process 300 terminates at block 352.
If the coordinator determines the OOB effective delay T at block 350OOBOr IB effective delay TIBNot expired, the process 300 returns to block 346.
After block 348, the process 300 terminates at block 352.
In the example process 300 described above, if the OOB success rate R is highOOB% and IB success rate RIB% high enough to reach the power capping target, only one power capping mode is used. In an implementation, the OOB power capping mode may be selected over the IB power capping mode because the OOB power capping mode does not occupy OS resources. Alternatively, IB power capping mode may be selected over OOB power capping mode. If the OOB success rate R is exceededOOB% and IB success rate RIB% who is not enough to reach the power capping target, then power capping is performed in both OOB power capping mode and IB power capping mode. During dual mode power capping, if any collisions are detected, the coordinator may unconditionally terminate one mode, e.g., IB power capping mode. Alternatively, the coordinator may terminate the OOB power capping mode. If both modes have been used and the power consumption of the computing system is still above the power capping target, the coordinator initiates a failover operational sequence flow. Further, the availability of cloud services may be improved.
Fig. 4 illustrates an example block diagram of a device 400 for implementing the processes and methods described above.
The device 400 includes one or more processors 402 and memory 404 communicatively coupled to the processors 402. Processor 402 executes one or more modules and/or processes to cause processor 402 to perform various functions. In implementations, the processor 402 may include a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), both a CPU and a GPU, or other processing units or components known in the art. Additionally, each processor 402 may have its own local memory, which may also store program modules, program data, and/or one or more operating systems. In an implementation, the memory 404 may be volatile, such as RAM, non-volatile, such as ROM, flash memory, a miniature hard drive, a memory card, etc., or some combination thereof.
The device 400 may additionally include an input/output (I/O) interface 406 for receiving and outputting data. The apparatus 400 may also include a communication module 408 that allows the apparatus 400 to communicate with other devices (not shown) over a network (not shown). The network may include the internet, wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, Radio Frequency (RF), infrared and other wireless media.
Memory 404 may include one or more computer-executable modules (modules) that may be executed by processor 402. In an implementation, the memory 404 may include, but is not limited to, a receiving module 410, a determining module 412, a power capping module 414, a waiting module 416, a power consumption determining module 418, a monitoring module 420, and a stopping module 422.
The receiving module 410 is configured to receive a power capping command that includes a power capping target of a computing system.
The determination module 412 is configured to determine whether the power consumption of the computing system meets a power capping target when performing power capping in one of the first power capping mode and the second power capping mode.
The determination module 412 is further configured to determine whether the first power contribution of the first power capping mode is greater than a first nominal success rate, perform power capping in the first power capping mode in response to determining that the first power contribution of the first power capping mode is greater than the first nominal success rate, wait until expiration of the first effective delay, and determine whether a power consumption of the computing system meets a power capping target when performing power capping in the first power capping mode.
The determination module 412 is further configured to determine whether the second power contribution of the second power capping mode is greater than the second nominal success rate, perform power capping in the second power capping mode in response to determining that the second power contribution of the second power capping mode is greater than the second nominal success rate, wait until expiration of the second effective delay, and determine whether a power consumption of the computing system meets a power capping target when performing power capping in the second power capping mode.
In an implementation, the first power capping mode is out-of-band (OOB) power capping and the second power capping mode is in-band (IB) power capping.
The power capping module 414 is configured to perform power capping in the first power capping mode and the second power capping mode in response to determining that the power consumption of the computing system does not meet the power capping target when performing power capping in one of the first power capping mode and the second power capping mode. For example, power capping may be performed in the first power capping mode and the second power capping mode simultaneously.
The wait module 416 is configured to wait until both the first effective delay and the second effective delay expire.
The power consumption determination module 418 is configured to determine whether the power consumption of the computing system meets a power capping target.
The monitoring module 420 is configured to monitor whether a collision occurs. The monitoring module 420 is further configured to check if a conflict occurs in the pcode/ucode setting.
The stopping module 422 is configured to stop performing power capping in one of the first power capping mode and the second power capping mode when the monitoring module detects the occurrence of the collision.
With the example apparatus 400 described above, the first power capping mode and the second power capping mode are coordinated and combined to achieve a power capping target for the computing system. Thus, power management during power capping is improved.
The processes and systems discussed herein may be implemented in, but are not limited to, distributed computing environments, parallel computing environments, clustered computing environments, grid computing environments, cloud computing environments, electric vehicles, power facilities, and the like.
Some or all of the operations of the above-described methods can be performed by executing computer readable instructions stored on a computer readable storage medium as defined below. The term "computer readable instructions" as used in the specification and claims includes routines, applications, application modules, program modules, programs, components, data structures, algorithms, and the like. Computer-readable instructions can be implemented on various system configurations, including single-processor or multiprocessor systems, minicomputers, mainframe computers, personal computers, hand-held computing devices, microprocessor-based, programmable consumer electronics, combinations thereof, and the like.
The computer-readable storage medium may include volatile memory (such as Random Access Memory (RAM)) and/or nonvolatile memory (such as Read Only Memory (ROM), flash memory, etc.). Computer-readable storage media may also include additional removable and/or non-removable storage devices, including, but not limited to, flash memory, magnetic storage, optical storage, and/or tape storage devices, which may provide non-volatile storage of computer-readable instructions, data structures, program modules, and the like.
Non-transitory computer-readable storage media are examples of computer-readable media. Computer-readable media includes at least two types of computer-readable media, namely computer-readable storage media and communication media. Computer-readable storage media includes volatile and nonvolatile, removable and non-removable media implemented in any process or technology for storage of information such as computer-readable instructions, data structures, program modules or other data. Computer-readable storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information for access by a computing device. In contrast, communication media may embody computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism. As defined herein, computer-readable storage media does not include communication media.
The computer-readable instructions stored on the one or more non-transitory computer-readable storage media, when executed by the one or more processors, may perform the operations described above with reference to fig. 1-4. Generally, computer readable instructions include routines, programs, objects, components, data structures, etc. that perform particular functions or implement particular abstract data types. The order in which the operations are described is not intended to be construed as a limitation, and any number of the described operations can be combined in any order and/or in parallel to implement the processes.
Example clauses
Clause 1. a method, comprising: receiving a power capping command, the power capping command comprising a power capping target of a computing system; determining whether a power consumption of the computing system meets the power capping target when performing the power capping in one of a first power capping mode and a second power capping mode; and performing the power capping in the first power capping mode and the second power capping mode in response to determining that the power consumption of the computing system does not satisfy the power capping target when performing the power capping in one of the first power capping mode and the second power capping mode.
Clause 2. the method of clause 1, wherein determining whether the power consumption of the computing system meets the power capping target when performing the power capping in one of the first power capping mode and the second power capping mode comprises: determining whether a first success rate of the first power capping mode is greater than a first nominal success rate; performing the power capping in the first power capping mode in response to determining that the first power contribution of the first power capping mode is greater than the first nominal success rate; waiting until expiration of the first effective delay; and determining whether the power consumption of the computing system meets the power capping target when performing the power capping in the first power capping mode.
Clause 3. the method of clause 1, wherein determining whether the power consumption of the computing system meets the power capping target when performing the power capping in one of the first power capping mode and the second power capping mode comprises: determining that a second power contribution of the second power capping mode is greater than a second nominal success rate; performing the power capping in the second power capping mode in response to determining that the second power contribution of the second power capping mode is greater than the second nominal success rate; waiting until expiration of a second effective delay; and determining whether the power consumption of the computing system meets the power capping target when performing the power capping in the second power capping mode.
Clause 4. the method of clause 1, further comprising: waiting until both the first effective delay and the second effective delay expire; and determining whether the power consumption of the computing system meets the power capping target.
Clause 5. the method of clause 1, further comprising: monitoring whether a conflict occurs; ceasing to perform the power capping in one of the first power capping mode and the second power capping mode in response to the monitoring of the occurrence of the collision.
Clause 6. the method of clause 5, wherein monitoring whether the conflict occurs comprises checking whether the conflict occurs in a pcode/ucode setting.
Clause 7. the method of clause 1, wherein the first power capping mode is out-of-band (OOB) power capping and the second power capping mode is in-band (IB) power capping.
Clause 8. a computer-readable storage medium storing computer-readable instructions executable by one or more processors, which, when executed by the one or more processors, cause the one or more processors to perform operations comprising: receiving a power capping command, the power capping command comprising a power capping target of a computing system; determining whether a power consumption of the computing system meets the power capping target when performing the power capping in one of a first power capping mode and a second power capping mode; and performing the power capping in the first power capping mode and the second power capping mode in response to determining that the power consumption of the computing system does not satisfy the power capping target when performing the power capping in one of the first power capping mode and the second power capping mode.
Clause 9. the computer-readable storage medium of clause 8, wherein determining whether the power consumption of the computing system meets the power capping target when performing the power capping in one of the first power capping mode and the second power capping mode comprises: determining whether a first success rate of the first power capping mode is greater than a first nominal success rate; performing the power capping in the first power capping mode in response to determining that the first power contribution of the first power capping mode is greater than the first nominal success rate; waiting until expiration of the first effective delay; and determining whether the power consumption of the computing system meets the power capping target when performing the power capping in the first power capping mode.
Clause 10 the computer-readable storage medium of clause 8, wherein determining whether the power consumption of the computing system meets the power capping target when performing the power capping in one of the first power capping mode and the second power capping mode comprises: determining whether a second power rating of the second power capping mode is greater than a second nominal success rate; performing the power capping in the second power capping mode in response to determining that the second power contribution of the second power capping mode is greater than the second nominal success rate; waiting until expiration of a second effective delay; and determining whether the power consumption of the computing system meets the power capping target when performing the power capping in the second power capping mode.
Clause 11. the computer-readable storage medium of clause 8, further comprising: waiting until both the first effective delay and the second effective delay expire; and determining whether the power consumption of the computing system meets the power capping target.
Clause 12. the computer-readable storage medium of clause 8, further comprising: monitoring whether a conflict occurs; ceasing to perform the power capping in one of the first power capping mode and the second power capping mode in response to the monitoring of the occurrence of the collision.
Clause 13. the computer-readable storage medium of clause 12, wherein monitoring whether the conflict occurs comprises checking whether the conflict occurs in a pcode/ucode setting.
Clause 14. the computer-readable storage medium of clause 8, wherein the first power capping mode is out-of-band (OOB) power capping and the second power capping mode is in-band (IB) power capping.
Clause 15. an apparatus, comprising: one or more processors; and a memory communicatively coupled to the one or more processors, the memory storing computer-executable modules executable by the one or more processors, the computer-executable modules comprising: a receiving module configured to receive a power capping command, the power capping command comprising a power capping target of a computing system; a determination module configured to determine whether a power consumption of the computing system meets the power capping target when performing the power capping in one of a first power capping mode and a second power capping mode; and a power capping module configured to perform the power capping in the first power capping mode and the second power capping mode in response to determining that the power consumption of the computing system does not meet the power capping target when performing the power capping in one of the first power capping mode and the second power capping mode.
Clause 16. the device of clause 15, wherein the determination module is further configured to: determining whether a first success rate of the first power capping mode is greater than a first nominal success rate; performing the power capping in the first power capping mode in response to determining that the first power contribution of the first power capping mode is greater than the first nominal success rate; waiting until expiration of the first effective delay; and determining whether the power consumption of the computing system meets the power capping target when performing the power capping in the first power capping mode.
Clause 17. the device of clause 15, wherein the determination module is further configured to: determining whether a second power rating of the second power capping mode is greater than a second nominal success rate; performing the power capping in the second power capping mode in response to determining that the second power contribution of the second power capping mode is greater than the second nominal success rate; waiting until expiration of a second effective delay; and determining whether the power consumption of the computing system meets the power capping target when performing the power capping in the second power capping mode.
Clause 18. the apparatus of clause 15, the computer-executable modules further comprising: a wait module configured to wait until both the first effective delay and the second effective delay expire; and a power consumption determination module configured to determine whether the power consumption of the computing system meets the power capping target.
Clause 19. the apparatus of clause 15, the computer-executable modules further comprising: a monitoring module configured to monitor whether a conflict occurs; a stopping module configured to stop performing the power capping in one of the first power capping mode and the second power capping mode when the monitoring module monitors the occurrence of the collision.
Clause 20. the device of clause 19, wherein the monitoring module is further configured to check whether the conflict occurs in a pcode/ucode setting.
Clause 21. the apparatus of clause 15, wherein the first power capping mode is out-of-band (OOB) power capping and the second power capping mode is in-band (IB) power capping.
Conclusion
Although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described. Rather, the specific features and acts are disclosed as example forms of implementing the claims.
Claims (21)
1. A method, the method comprising:
receiving a power capping command, the power capping command comprising a power capping target of a computing system;
determining whether a power consumption of the computing system meets the power capping target when performing the power capping in one of a first power capping mode and a second power capping mode; and
performing the power capping in the first power capping mode and the second power capping mode in response to determining that the power consumption of the computing system does not satisfy the power capping target when performing the power capping in one of the first power capping mode and the second power capping mode.
2. The method of claim 1, wherein determining whether the power consumption of the computing system meets the power capping target when performing the power capping in one of the first power capping mode and the second power capping mode comprises:
determining whether a first success rate of the first power capping mode is greater than a first nominal success rate;
performing the power capping in the first power capping mode in response to determining that the first power contribution of the first power capping mode is greater than the first nominal success rate;
waiting until expiration of the first effective delay; and
determining whether the power consumption of the computing system meets the power capping target when performing the power capping in the first power capping mode.
3. The method of claim 1, wherein determining whether the power consumption of the computing system meets the power capping target when performing the power capping in one of the first power capping mode and the second power capping mode comprises:
determining that a second power contribution of the second power capping mode is greater than a second nominal success rate;
performing the power capping in the second power capping mode in response to determining that the second power contribution of the second power capping mode is greater than the second nominal success rate;
waiting until expiration of a second effective delay; and
determining whether the power consumption of the computing system meets the power capping target when performing the power capping in the second power capping mode.
4. The method of claim 1, further comprising:
waiting until both the first effective delay and the second effective delay expire; and
determining whether the power consumption of the computing system satisfies the power capping target.
5. The method of claim 1, further comprising:
monitoring whether a conflict occurs;
ceasing to perform the power capping in one of the first power capping mode and the second power capping mode in response to the monitoring of the occurrence of the collision.
6. The method of claim 5, wherein monitoring whether the conflict occurs comprises checking whether the conflict occurs in a pcode/ucode setting.
7. The method of claim 1, wherein the first power capping mode is out-of-band (OOB) power capping and the second power capping mode is in-band (IB) power capping.
8. A computer-readable storage medium storing computer-readable instructions executable by one or more processors, the computer-readable instructions, when executed by the one or more processors, causing the one or more processors to perform operations comprising:
receiving a power capping command, the power capping command comprising a power capping target of a computing system;
determining whether a power consumption of the computing system meets the power capping target when performing the power capping in one of a first power capping mode and a second power capping mode; and
performing the power capping in the first power capping mode and the second power capping mode in response to determining that the power consumption of the computing system does not satisfy the power capping target when performing the power capping in one of the first power capping mode and the second power capping mode.
9. The computer-readable storage medium of claim 8, wherein determining whether the power consumption of the computing system satisfies the power capping target when performing the power capping in one of the first power capping mode and the second power capping mode comprises:
determining whether a first success rate of the first power capping mode is greater than a first nominal success rate;
performing the power capping in the first power capping mode in response to determining that the first power contribution of the first power capping mode is greater than the first nominal success rate;
waiting until expiration of the first effective delay; and
determining whether the power consumption of the computing system meets the power capping target when performing the power capping in the first power capping mode.
10. The computer-readable storage medium of claim 8, wherein determining whether the power consumption of the computing system satisfies the power capping target when performing the power capping in one of the first power capping mode and the second power capping mode comprises:
determining whether a second power rating of the second power capping mode is greater than a second nominal success rate;
performing the power capping in the second power capping mode in response to determining that the second power contribution of the second power capping mode is greater than the second nominal success rate;
waiting until expiration of a second effective delay; and
determining whether the power consumption of the computing system meets the power capping target when performing the power capping in the second power capping mode.
11. The computer-readable storage medium of claim 8, further comprising:
waiting until both the first effective delay and the second effective delay expire; and
determining whether the power consumption of the computing system satisfies the power capping target.
12. The computer-readable storage medium of claim 8, further comprising:
monitoring whether a conflict occurs;
ceasing to perform the power capping in one of the first power capping mode and the second power capping mode in response to the monitoring of the occurrence of the collision.
13. The computer-readable storage medium of claim 12, wherein monitoring whether the conflict occurs comprises checking whether the conflict occurs in a pcode/ucode setting.
14. The computer-readable storage medium of claim 8, wherein the first power capping mode is out-of-band (OOB) power capping and the second power capping mode is in-band (IB) power capping.
15. An apparatus, the apparatus comprising:
one or more processors; and
a memory communicatively coupled to the one or more processors, the memory storing computer-executable modules executable by the one or more processors, the computer-executable modules comprising:
a receiving module configured to receive a power capping command, the power capping command comprising a power capping target of a computing system;
a determination module configured to determine whether a power consumption of the computing system meets the power capping target when performing the power capping in one of a first power capping mode and a second power capping mode; and
a power capping module configured to perform the power capping in the first power capping mode and the second power capping mode in response to determining that the power consumption of the computing system does not meet the power capping target when performing the power capping in one of the first power capping mode and the second power capping mode.
16. The device of claim 15, wherein the determination module is further configured to:
determining whether a first success rate of the first power capping mode is greater than a first nominal success rate;
performing the power capping in the first power capping mode in response to determining that the first power contribution of the first power capping mode is greater than the first nominal success rate;
waiting until expiration of the first effective delay; and
determining whether the power consumption of the computing system meets the power capping target when performing the power capping in the first power capping mode.
17. The device of claim 15, wherein the determination module is further configured to:
determining whether a second power rating of the second power capping mode is greater than a second nominal success rate;
performing the power capping in the second power capping mode in response to determining that the second power contribution of the second power capping mode is greater than the second nominal success rate;
waiting until expiration of a second effective delay; and
determining whether the power consumption of the computing system meets the power capping target when performing the power capping in the second power capping mode.
18. The apparatus of claim 15, the computer-executable modules further comprising:
a wait module configured to wait until both the first effective delay and the second effective delay expire; and
a power consumption determination module configured to determine whether the power consumption of the computing system satisfies the power capping target.
19. The apparatus of claim 15, the computer-executable modules further comprising:
a monitoring module configured to monitor whether a conflict occurs;
a stopping module configured to stop performing the power capping in one of the first power capping mode and the second power capping mode when the monitoring module monitors the occurrence of the collision.
20. The apparatus of claim 19, wherein the monitoring module is further configured to check whether the conflict occurs in a pcode/ucode setting.
21. The apparatus of claim 15, wherein the first power capping mode is out-of-band (OOB) power capping and the second power capping mode is in-band (IB) power capping.
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2019/108555 WO2021056416A1 (en) | 2019-09-27 | 2019-09-27 | Power management method and apparatus |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114402272A true CN114402272A (en) | 2022-04-26 |
| CN114402272B CN114402272B (en) | 2024-03-19 |
Family
ID=75165033
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201980099694.XA Active CN114402272B (en) | 2019-09-27 | 2019-09-27 | Power management method and apparatus |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN114402272B (en) |
| WO (1) | WO2021056416A1 (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030196126A1 (en) * | 2002-04-11 | 2003-10-16 | Fung Henry T. | System, method, and architecture for dynamic server power management and dynamic workload management for multi-server environment |
| CN101183917A (en) * | 2007-12-24 | 2008-05-21 | 腾讯科技（深圳）有限公司 | Multimedia data coding velocity regulating method and device |
| US20090150693A1 (en) * | 2007-12-05 | 2009-06-11 | Vivek Kashyap | Method for power capping with co-operative dynamic voltage and frequency scaling |
| CN101477403A (en) * | 2009-01-22 | 2009-07-08 | 浪潮电子信息产业股份有限公司 | Automatic control method for system power consumption |
| US20110035611A1 (en) * | 2009-08-07 | 2011-02-10 | International Business Machines Corporation | Coordinating in-band and out-of-band power management |
| US20170336855A1 (en) * | 2016-05-20 | 2017-11-23 | Dell Products L.P. | Systems and methods for chassis-level view of information handling system power capping |
| CN108983946A (en) * | 2018-06-13 | 2018-12-11 | 烽火通信科技股份有限公司 | A kind of server energy consumption control method, system and equipment |
-
2019
- 2019-09-27 CN CN201980099694.XA patent/CN114402272B/en active Active
- 2019-09-27 WO PCT/CN2019/108555 patent/WO2021056416A1/en active Application Filing
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030196126A1 (en) * | 2002-04-11 | 2003-10-16 | Fung Henry T. | System, method, and architecture for dynamic server power management and dynamic workload management for multi-server environment |
| US20090150693A1 (en) * | 2007-12-05 | 2009-06-11 | Vivek Kashyap | Method for power capping with co-operative dynamic voltage and frequency scaling |
| CN101183917A (en) * | 2007-12-24 | 2008-05-21 | 腾讯科技（深圳）有限公司 | Multimedia data coding velocity regulating method and device |
| CN101477403A (en) * | 2009-01-22 | 2009-07-08 | 浪潮电子信息产业股份有限公司 | Automatic control method for system power consumption |
| US20110035611A1 (en) * | 2009-08-07 | 2011-02-10 | International Business Machines Corporation | Coordinating in-band and out-of-band power management |
| US20170336855A1 (en) * | 2016-05-20 | 2017-11-23 | Dell Products L.P. | Systems and methods for chassis-level view of information handling system power capping |
| CN108983946A (en) * | 2018-06-13 | 2018-12-11 | 烽火通信科技股份有限公司 | A kind of server energy consumption control method, system and equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2021056416A1 (en) | 2021-04-01 |
| CN114402272B (en) | 2024-03-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20190258506A1 (en) | Systems and methods of host-aware resource management involving cluster-based resource pools | |
| AU2011299337B2 (en) | Controlled automatic healing of data-center services | |
| US20160378570A1 (en) | Techniques for Offloading Computational Tasks between Nodes | |
| US9614779B2 (en) | Cloud compute scheduling using a heuristic contention model | |
| US20070028239A1 (en) | Dynamic performance management for virtual servers | |
| US9547359B2 (en) | Dynamic system management communication path selection | |
| US9513690B2 (en) | Apparatus and method for adjusting operating frequencies of processors based on result of comparison of power level with a first threshold and a second threshold | |
| US8810584B2 (en) | Smart power management in graphics processing unit (GPU) based cluster computing during predictably occurring idle time | |
| US20210119878A1 (en) | Detection and remediation of virtual environment performance issues | |
| US9952911B2 (en) | Dynamically optimized device driver protocol assist threads | |
| US20190384667A1 (en) | System and method for creating and using hybrid virtual computing systems | |
| Baccarelli et al. | Minimum-energy bandwidth management for QoS live migration of virtual machines | |
| US20140189173A1 (en) | Network Efficiency and Power Savings | |
| US9250668B2 (en) | Decoupled power and performance allocation in a multiprocessing system | |
| US10114438B2 (en) | Dynamic power budgeting in a chassis | |
| US9712647B2 (en) | Optimization of computing resources through monitoring and manipulating availabilty | |
| US11243603B2 (en) | Power management of an event-based processing system | |
| US10996942B1 (en) | System and method for graphics processing unit firmware updates | |
| CN114402272B (en) | Power management method and apparatus | |
| US9501128B2 (en) | Cooperative reduced power mode suspension for high input/output (‘I/O’) workloads | |
| US8312126B2 (en) | Managing at least one computer node | |
| CN116339902A (en) | Event message management in a super converged infrastructure environment | |
| US10379928B2 (en) | Preventing software component timer processing issues | |
| US11782490B2 (en) | Software-defined fail-safe power draw control for rack power distribution units | |
| US20220271555A1 (en) | Information handling system with battery power management |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |