CN116755542B - Whole machine power consumption reduction method, system, substrate management controller and server - Google Patents

Whole machine power consumption reduction method, system, substrate management controller and server Download PDF

Info

Publication number
CN116755542B
CN116755542B CN202310715603.9A CN202310715603A CN116755542B CN 116755542 B CN116755542 B CN 116755542B CN 202310715603 A CN202310715603 A CN 202310715603A CN 116755542 B CN116755542 B CN 116755542B
Authority
CN
China
Prior art keywords
power consumption
module
power supply
management controller
cpu
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.)
Active
Application number
CN202310715603.9A
Other languages
Chinese (zh)
Other versions
CN116755542A (en
Inventor
申明伟
马振鹏
茅振宇
杨占
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hexin Technology Co ltd
Beijing Hexin Digital Technology Co ltd
Original Assignee
Hexin Technology Co ltd
Beijing Hexin Digital Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Hexin Technology Co ltd, Beijing Hexin Digital Technology Co ltd filed Critical Hexin Technology Co ltd
Priority to CN202310715603.9A priority Critical patent/CN116755542B/en
Publication of CN116755542A publication Critical patent/CN116755542A/en
Application granted granted Critical
Publication of CN116755542B publication Critical patent/CN116755542B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Landscapes

  • Power Sources (AREA)

Abstract

The invention provides a method and a system for reducing overall power consumption, a baseboard management controller and a server, wherein the method comprises the following steps: responding to the starting operation of the server system, setting the operation state of the system into a normal operation mode through a baseboard management controller, setting each component of the system into corresponding normal operation power consumption, and monitoring and triggering a power consumption reduction event in real time; when the baseboard management controller monitors the trigger power consumption reduction event, the running state of the system is adjusted to be in a low-power consumption working mode, and after a preset time length, the power consumption of the system is dynamically adjusted according to the power supply capacity of the power supply until the system receives a shutdown command to shutdown. The invention not only can effectively cope with the power failure scene, quickly reduce the power consumption under the condition of ensuring the normal operation of the server system, avoid the system downtime caused by PSU overload, effectively maintain the stability of the system and the continuity of the service, but also can meet the power consumption dynamic adjustment requirement of the network terminal, and provide convenience for more operation and maintenance adjustment and measurement scenes.

Description

Whole machine power consumption reduction method, system, substrate management controller and server
Technical Field
The present invention relates to the field of computer technologies, and in particular, to a method and a system for reducing power consumption of a complete machine, a baseboard management controller, and a server.
Background
The server system often adopts CRPS (Common Redundant Power Supplies) power supply modules to supply power, and because CRPS power supply modules have limited power supply capability, when the power consumption of the whole system is greater than the power supply capability range of the CRPS power supply modules, at least two power supplies are required to be configured, or the CRPS power supply adopts a double-loop power supply to supply power, one half of the CRPS power supply adopts mains supply to supply power, and the other half of the CRPS power supply adopts UPS (Uninterruptible Power System) to supply power. However, when one or more of the power supplies is damaged or one of the circuits fails, the total output capacity of the CRPS power supply is rapidly reduced, and at this time, the power consumption of the system is far greater than the power supply capacity of the CRPS power supply, so that the CRPS power supply is overloaded, and the system is turned off accidentally.
Therefore, based on the current state that the CRPS power supply allows overload output to be only tens of milliseconds, it is needed to provide a method for reducing overall power consumption of a complete machine, which can reasonably reduce overall power consumption of the system in a short time and avoid unexpected shutdown risk to the greatest extent.
Disclosure of Invention
The invention aims to provide a method for reducing the power consumption of a whole machine, which monitors a trigger power consumption reduction event in real time through a baseboard management controller, dynamically adjusts the power consumption of a system according to the power supply capacity of a power supply after the corresponding event is monitored, solves the application defect that the power failure of the existing server is extremely easy to cause the unexpected shutdown of the system, can effectively cope with a power failure scene, quickly reduces the power consumption under the condition of ensuring the normal operation of the server system, avoids the system downtime caused by PSU overload, and effectively maintains the stability of the system and the continuity of service.
In order to achieve the above object, it is necessary to provide a method, a system, a baseboard management controller and a server for reducing power consumption of a whole machine.
In a first aspect, an embodiment of the present invention provides a method for reducing power consumption of a complete machine, where the method includes the following steps:
Responding to the starting operation of the server system, setting the operation state of the system into a normal operation mode through a baseboard management controller, setting each component of the system into corresponding normal operation power consumption, and monitoring and triggering a power consumption reduction event in real time;
when the baseboard management controller monitors the trigger power consumption reduction event, the system running state is adjusted to be in a low power consumption working mode, and after a preset time period, the system power consumption is dynamically adjusted according to the power supply capacity of the power supply until the system receives a shutdown command to shutdown.
Further, the triggering power consumption reduction event includes a power failure, a power supply failure, or the network terminal sending a power consumption reduction request.
Further, the step of dynamically adjusting the system power consumption according to the power supply capability of the power supply includes:
According to the power supply capacity of the power supply, carrying out power consumption reduction adjustment on the pre-screened power consumption module to be reduced; the pre-screened power consumption module to be reduced comprises a CPU module and a fan module.
Further, the step of performing power consumption reduction adjustment on the pre-screened power consumption module to be reduced according to the power supply capability of the power supply includes:
And transmitting a frequency-reducing command to a CPU module interface through the baseboard management controller, and reducing the duty ratio of a rotating speed signal corresponding to the fan module.
Further, the step of transmitting the down-conversion command to the CPU module interface through the baseboard management controller includes:
Reducing cTDP register values corresponding to the CPU module to a first preset value; or, pulling down the CPU down-conversion signal to a second preset value through the general input/output port;
acquiring power output power consumption and current system power consumption in real time, and judging whether the power output power consumption is larger than the current system power consumption or not;
If yes, the cTDP register value corresponding to the CPU module is increased to the maximum allowable CPU frequency value, otherwise, the cTDP register value corresponding to the CPU module is kept unchanged.
Further, the step of reducing, by the baseboard management controller, a duty ratio of a rotational speed signal corresponding to the fan module includes:
Reducing pulse width modulation registers of all fan rotating speed controllers corresponding to the fan modules to a third preset value;
shortening the system temperature polling interval, and acquiring the CPU polling temperature and the memory polling temperature of the system through the baseboard management controller;
Judging whether a preset working condition is met or not according to the CPU polling temperature and the memory polling temperature; the preset working condition is that the CPU polling temperature is within a first preset temperature threshold value, and the memory polling temperature is within a second preset temperature threshold value;
If not, the duty ratio of the rotating speed signal corresponding to the fan module is gradually increased according to a preset speed increasing step length until the CPU polling temperature and the memory polling temperature meet the preset working conditions, and after the CPU polling temperature and the memory polling temperature meet the preset working conditions, the pulse width modulation register of each fan rotating speed controller corresponding to the fan module is reduced to a third preset value.
Further, the step of dynamically adjusting the system power consumption according to the power supply capability of the power supply further includes:
When the system running state is in a low-power-consumption working mode, monitoring and recovering normal power consumption events in real time through the baseboard management controller;
When the baseboard management controller monitors the normal power consumption recovery event, the system running state is adjusted to a normal working mode, the CPU module and the fan module are adjusted to corresponding normal running power consumption, and the system temperature polling interval is recovered to a normal temperature polling interval.
In a second aspect, an embodiment of the present invention provides a system for reducing power consumption of a complete machine, where the system includes:
The initialization module is used for responding to the starting operation of the server system, setting the operation state of the system into a normal operation mode through the baseboard management controller, setting each part of the system into corresponding normal operation power consumption, and monitoring and triggering a power consumption reduction event in real time;
And the power consumption adjusting module is used for adjusting the running state of the system into a low-power consumption working mode when the baseboard management controller monitors the trigger power consumption reducing event, and dynamically adjusting the power consumption of the system according to the power supply capacity of the power supply after a preset time period until the system is shut down.
In a third aspect, an embodiment of the present invention further provides a baseboard management controller, including a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor implements the steps of the method when executing the computer program.
In a fourth aspect, an embodiment of the present invention further provides a server, including:
The substrate management controller is connected with the CPU module, the fan module, the memory module, the hard disk module and the PCIe card.
The application provides a method, a system, a baseboard management controller and a server for reducing the power consumption of the whole machine, by the method, the technical scheme that the system power consumption is dynamically adjusted according to the power supply capacity until the system is shut down after a preset time length is reached, the system running state is set to a normal working mode by the baseboard management controller, each part of the system is set to corresponding normal running power consumption, a trigger power consumption reduction event is monitored in real time, and the system running state is adjusted to a low power consumption working mode when the baseboard management controller monitors the trigger power consumption reduction event. Compared with the prior art, the method for reducing the power consumption of the whole machine can effectively cope with a power failure scene, can rapidly reduce the power consumption under the condition of ensuring the normal operation of a server system, avoids the system downtime caused by PSU overload, effectively maintains the stability of the system and the continuity of service, can meet the power consumption dynamic adjustment requirement of a network terminal, provides convenience for more operation and maintenance adjustment scenes, and has higher application value.
Drawings
FIG. 1 is a schematic diagram of a method for reducing power consumption of a complete machine according to an embodiment of the present invention;
FIG. 2 is a flow chart of a method for reducing power consumption of a complete machine according to an embodiment of the invention;
FIG. 3 is a schematic diagram of the power consumption ratio of each functional module of the general 2u dual-path server according to an embodiment of the present invention;
FIG. 4 is a detailed flow chart of dynamic adjustment of system power consumption according to the power supply capability in the embodiment of the invention;
fig. 5 is a system power consumption trend chart of the embodiment of the invention, wherein the system power consumption trend chart is used for executing the method for reducing the power consumption of the whole machine when the BMC monitors the power failure;
Fig. 6 is a schematic structural diagram of a system for reducing power consumption of a complete machine in an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantageous effects of the present application more apparent, the present application will be further described in detail with reference to the accompanying drawings and examples, and it is apparent that the examples described below are part of the examples of the present application, which are provided for illustration only and are not intended to limit the scope of the present application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
The method for reducing the power consumption of the whole machine is a method for monitoring various triggering power consumption reduction events based on a baseboard management controller (Baseboard Management Controller, BMC) which is realized on the basis of the existing flow architecture shown in the figure 1, timely adjusting the working state of a system and carrying out efficient and reliable dynamic regulation and control on a power consumption module of the system according to the power supply capacity of the system, and can not only ensure the normal operation of the system when facing power failure or other scenes needing to reduce the power consumption of the system, but also quickly reduce the power consumption of the system, avoid the system downtime caused by PSU overload, effectively maintain the stability of the system and the continuity of service, and quickly restore the server system to the normal state when corresponding failure recovery or other scenes needing to restore the normal power consumption. The following embodiments will describe the overall power consumption reduction method of the present invention in detail.
In one embodiment, as shown in fig. 2, a method for reducing power consumption of a complete machine is provided, including the following steps:
s11, responding to the starting-up operation of the server system, setting the operation state of the system into a normal operation mode through a baseboard management controller, setting each part of the system into corresponding normal operation power consumption, and monitoring and triggering a power consumption reduction event in real time; the server system can be divided into a CPU module, a fan module, a memory module, a hard disk module, a PCIe card and the like according to functional modules, and a baseboard management controller BMC which is connected with the functional modules and is used for carrying out real-time monitoring management on the functional modules; the system running state in this embodiment may be understood as a working mode determined based on whether the current system is in a normal running power consumption state, including a normal working mode and a low power consumption working mode; the normal working mode can be understood as that the BMC does not monitor the trigger power consumption reduction event, and each functional module of the system operates according to the established normal power consumption. In principle, the maintenance and update of the system running state can be realized by adding other functional modules, but considering that the BMC can monitor the system state in real time in the running process of the server system, and the power consumption adjustment needs to depend on the relevant monitoring result of the BMC, in order to realize simple and efficient maintenance and update of the system running state under the condition of reducing the system variation as much as possible and not introducing the system running risk, the embodiment is preferable, the BMC directly maintains the system running state, namely the BMC defaults the system running state to be in a normal working mode when the server is normally started, updates and maintains the system running state according to the condition of triggering the power consumption reduction event by real-time monitoring, and provides a reliable basis for the subsequent efficient and accurate dynamic adjustment of the system power consumption;
The above-mentioned triggering power consumption reduction event may in principle only include the situation that the power supply module (PC Power supply unit, PSU) is accidentally powered down (out of place), or the power supply capability of the power supply module is failed (in place but the power supply capability is insufficient), that is, only the BMC is required to monitor whether the PSU is normal (in place) and whether the PSU is normal (whether the power supply capability is normal) in working state, but in consideration of the actual application process of the server system, certain operation and maintenance adjustment scenes may be involved, the system power consumption needs to be rapidly reduced to perform problem location or performance verification under the situation that the power supply is in place and the power supply capability is normal, in order to effectively meet the power consumption dynamic adjustment requirement of the network terminal, convenience is provided for more operation and maintenance adjustment scenes, and in this embodiment, the triggering power consumption reduction event is preferably expanded; specifically, the triggering power consumption reduction event includes a power failure, a power supply failure or a network terminal sending a power consumption reduction request, where the network terminal sending the power consumption reduction request may be understood as a request for quickly reducing power consumption, which is sent by a relevant operation and maintenance person or a machine room control end through a network and received and processed by the BMC;
S12, when the baseboard management controller monitors the trigger power consumption reduction event, the running state of the system is adjusted to be in a low-power consumption working mode, and after a preset time period, the power consumption of the system is dynamically adjusted according to the power supply capacity of the power supply until the system receives a shutdown command to shutdown; the low-power-consumption working mode can be understood as a system running state that a part of functional modules of the system corresponding to the normal working mode run in a power consumption state lower than the normal power consumption; when the server system is in normal operation in the system operation state of the normal working mode, if the baseboard management controller monitors any trigger event of power failure, power supply failure or power consumption reduction request sent by the network terminal, the system operation state is updated to the low-power consumption working mode at the first time so as to ensure the efficient execution of the subsequent system power consumption adjustment flow; meanwhile, when the BMC updates the running state of the system into a low-power-consumption working mode, the running state of the system can be synchronously entered into a system power consumption adjustment flow in principle for quickly reducing power consumption, but in consideration of the fact that certain monitoring results may be misaligned in the actual running process, in order to ensure that the system power consumption is effectively adjusted in time when the monitoring results are actually required to be reduced, unnecessary adjustment or twice reverse adjustment in a short time can be effectively avoided, which is caused by the fact that the monitoring results are misaligned, so that the risk of running instability of the server system is caused, the embodiment preferably carries out dynamic adjustment on the system power consumption according to the power supply capacity after the system enters the low-power-consumption working mode for a period of time, controls the whole power consumption within a safe range, and avoids the system downtime caused by overload protection and stopping output of PSU; it should be noted that the preset duration may be selected according to the actual application requirement, which is not limited herein.
In addition, after the server starts to run, no matter whether the system has load or not, the functional modules such as the CPU module, the fan module, the memory module, the hard disk module, the PCIe card and the like have corresponding power consumption, so when the functional modules are selected as power consumption adjustment objects in the situation of fast lower system power consumption is critical to the overall power consumption reduction operation. In order to ensure that the selection process of each power consumption adjustment object is simple and reliable, and the application requirement of effectively reducing the power consumption of the whole system in a short time can be met, the embodiment is preferable in that the power consumption of each functional module in the server is subjected to statistical analysis in advance, the functional module with larger maximum power consumption and controllable power consumption adjustment influence is screened out to be used as the power consumption module to be reduced, and when the power consumption needs to be reduced, the corresponding adjustment operation can be directly carried out on each power consumption module to be reduced; specifically, the step of dynamically adjusting the system power consumption according to the power supply capability of the power supply includes:
According to the power supply capacity of the power supply, carrying out power consumption reduction adjustment on the pre-screened power consumption module to be reduced; the pre-screened power consumption module to be reduced comprises a CPU module and a fan module; the determination process of the power consumption module to be reduced is described by the following general two-way server example:
acquiring power consumption information of each functional module of the server under the two conditions of no load and load of the system through a BMC web page in advance, summarizing the acquired power consumption information under the two conditions to obtain a module statistical result shown in a rear table 1, and analyzing based on the table 1 to obtain the power consumption ratio of each functional module shown in fig. 3;
Table 1 individual component power consumption information for each functional module of a server system
Component classification Individual component power consumption Quantity of Power consumption Power consumption duty cycle
Processor and method for controlling the same 250 2 500 0.242718447
Memory 15 32 480 0.233009709
Hard disk 15 12 180 0.087378641
PCIe card 75 6 450 0.218446602
Fan with fan body 75 6 450 0.218446602
Totals to 2060
Based on the analysis result shown in fig. 3, the power consumption of the processor (CPU), the fan, the memory and the PCIe card are more than twenty percent on average, but during the operation of the server, the modules such as the memory, the hard disk, the PCIe card and the like cannot be directly turned off, otherwise, the problems such as hardware resource loss and downtime of the operating system are caused, the power consumption of the remaining CPU module and the fan module can be adjusted, and as can be seen from the duty ratio, the total power consumption of the CPU module and the fan module is close to 50%, if the power consumption close to 50% can be reduced, the power PSU cannot be overloaded, and the normal operation of the system can be maintained; therefore, the purpose of rapid power consumption reduction can be achieved by adjusting the power consumption of the CPU module and the fan module, namely, the CPU module and the fan module can be directly used as the power consumption module to be reduced, and when the BMC monitors power failure, power supply failure or a network terminal sends a power consumption reduction request and other trigger power consumption reduction events, the CPU module and the fan module can be directly subjected to corresponding power consumption adjustment;
in the embodiment, through carrying out statistical analysis on the power consumption information of each functional module in the actual running of the server and combining with the actual working principle of the server, the to-be-reduced power consumption module which can be directly used for adjustment after the BMC monitors the trigger power consumption reduction event is screened out, so that the rationality and reliability of selection of the power consumption adjustment module are ensured, and when the power consumption reduction processing is determined to be required, the selected to-be-reduced power consumption module can be directly and timely adjusted in a preset mode, the operation duration of reducing the system power consumption can be effectively shortened, and further, the guarantee is provided for rapid reduction of the system power consumption;
In addition, in order to reduce system variation as much as possible, without introducing additional system operation risk and reducing system development and operation cost, the embodiment preferably adopts a substrate management controller to transmit corresponding power consumption reduction instructions to a CPU module and a fan module to realize efficient and reliable power consumption reduction adjustment of the power consumption module to be reduced; specifically, as shown in fig. 4, the step of performing power consumption reduction adjustment on the pre-screened power consumption module to be reduced according to the power supply capability of the power supply includes:
transmitting a frequency-reducing command to a CPU module interface through the baseboard management controller, and reducing the duty ratio of a rotating speed signal corresponding to the fan module;
the down-conversion command can be transmitted to the CPU module interface in principle by directly writing cTDP (Configable THERMAL DESIGN Power) register values, but in order to ensure the execution effect of reducing the Power consumption of the CPU module, the embodiment preferably provides another down-conversion mode at the same time so as to meet the selection requirements of different application scenarios; specifically, the step of transmitting the down-conversion command to the CPU module interface through the baseboard management controller includes:
Reducing cTDP register values corresponding to the CPU module to a first preset value; the first preset value may be understood as a minimum frequency value which is determined according to an actual application scenario and can be written into cTDP register values by the BMC and ensures the operation of the CPU, and is not particularly limited herein;
Or, pulling down the CPU down-conversion signal to a second preset value through the general input/output port; the second preset value may also be understood as a minimum CPU thread signal value that is determined according to an actual application scenario and that can be input by the BMC through the GPIO and that ensures the operation of the CPU, which is not specifically limited herein.
After the operation frequency of the CPU is reduced through the steps, the effective reduction of the corresponding operation power consumption can be synchronously realized, namely, if the corresponding first preset value or second preset value is to reduce the operation frequency of the CPU module to the minimum value, the corresponding operation power consumption is also reduced to the minimum value; meanwhile, as the running frequency of the CPU module is reduced, the operation of the CPU module on components such as a memory module, a hard disk module, a PCIe card and the like is also reduced, and the power consumption of the corresponding components is naturally and synchronously reduced; in principle, the method can realize the efficient and reliable reduction of the running power consumption of the CPU module, but considers that the power supply capacity (output power consumption) of the system power supply PSU can exceed the adjusted system power consumption requirement due to the fact that the power consumption adjustment threshold value is used, so that the utilization rate condition of the output power consumption of the power supply PSU is reduced, and in order to ensure that the CPU performance is maximized within the power supply capacity range of the power supply PSU and further ensure the processing efficiency of the server system, in the embodiment, after the power consumption of the CPU module is reduced, whether the frequency of the CPU module needs to be finely adjusted is further judged according to the adjusted power supply output power consumption and the current system power consumption; specifically, the step of transmitting the down-conversion command to the CPU module interface through the baseboard management controller further includes:
acquiring power output power consumption and current system power consumption in real time, and judging whether the power output power consumption is larger than the current system power consumption or not;
If yes, the cTDP register value corresponding to the CPU module is increased to the maximum allowable frequency value of the current CPU, otherwise, the cTDP register value corresponding to the CPU module is kept unchanged; the maximum allowable frequency value of the current CPU can be understood as the maximum operating frequency of the CPU which can be supported in the current power output power consumption range, and can be determined according to the actual system operating condition, and the maximum allowable frequency value is not particularly limited herein; it should be noted that, after the adjustment processing is performed on the cTDP register values by acquiring the power output power consumption and the current system power consumption in real time, the CPU operating frequency needs to be dynamically adjusted by repeatedly acquiring the power output power consumption and the current system power consumption to perform judgment and analysis, so as to ensure the rationality of the system power consumption adjustment at different moments as much as possible.
Specifically, the step of reducing, by the baseboard management controller, a duty ratio of a rotational speed signal corresponding to the fan module includes:
reducing pulse width modulation registers of all fan rotating speed controllers corresponding to the fan modules to a third preset value; the third threshold value may be understood as an expected fan speed PWM signal duty cycle value that may be determined by the BMC according to an actual application scenario and expected to be reached by writing PWM AND FAN Tachometer Controller related registers, and the fan speed PWM signal duty cycle may be adjusted to be 10 or lower, which is not particularly limited herein; for example, if there are 6 fan units in the fan module, the BMC adjusts the duty ratio of the PWM signal of the output fan speed by controlling the fan speed, so as to rapidly reduce the power consumption of the fan module, as follows:
Rotational speed control signal of fan 1: writing a register 0x1E61:0000+004H to realize FAN1PWM duty cycle data control, thereby realizing the rotation speed control of the FAN 1;
rotational speed control signal of fan 2: writing a register 0x1E61:0000+010H to realize the data control of the FAN2PWM duty ratio, thereby realizing the rotation speed control of the FAN 2;
rotational speed control signal of fan 3: writing a register 0x1E61:0000+01CH to realize the data control of the FAN3PWM duty ratio, thereby realizing the rotation speed control of the FAN 3;
Rotational speed control signal of fan 4: writing a register 0x1E61:0000+028H to realize the control of FAN4PWM duty cycle data, thereby realizing the control of the rotating speed of the FAN 4;
rotational speed control signal of fan 5: writing a register 0x1E61:0000+034H to realize the data control of the FAN5PWM duty cycle, thereby realizing the rotation speed control of the FAN 5;
rotational speed control signal of fan 5: the write register 0x1E61:0000+040H realizes the data control of the FAN6PWM duty cycle, thereby realizing the rotation speed control of the FAN 6.
In practical application, the method can realize the rapid reduction of the running power consumption of the fan module, in principle, the running power consumption of the modules such as a CPU, a memory and a hard disk can be synchronously reduced due to the reduction of the running frequency of the CPU, the corresponding temperature rise is slower, the synchronous reduction of the fan rotating speed has little influence on the system, but the stability of a server and the continuous requirement of service are considered, in order to avoid the risk of system downtime caused by the over-temperature of the server system due to the too fast temperature rise of the CPU module and the memory module under the condition of reducing the fan rotating speed as far as possible, the implementation preferably shortens the temperature polling interval of the system after the reduction of the power consumption of the fan module, and strengthens the temperature monitoring of the CPU and the memory so as to discover the temperature abnormality in time for necessary regulation; specifically, the step of reducing, by the baseboard management controller, the duty ratio of the rotational speed signal corresponding to the fan module further includes:
Shortening the system temperature polling interval, and acquiring the CPU polling temperature and the memory polling temperature of the system through the baseboard management controller; the shortening degree of the system temperature polling interval can be set according to practical application requirements, and is not particularly limited herein;
Judging whether a preset working condition is met or not according to the CPU polling temperature and the memory polling temperature; the preset working condition is that the CPU polling temperature is within a first preset temperature threshold value, and the memory polling temperature is within a second preset temperature threshold value; the first preset temperature threshold and the second preset temperature threshold can be understood as corresponding protection temperature threshold points allowed by system operation, can be selected according to actual application requirements, can be set identically or differently, and are not particularly limited herein;
If not, gradually increasing the duty ratio of the rotating speed signal corresponding to the fan module according to a preset speed increasing step length until the CPU polling temperature and the memory polling temperature meet the preset working conditions, and reducing the pulse width modulation register of each fan rotating speed controller corresponding to the fan module to a third preset value after the CPU polling temperature and the memory polling temperature meet the preset working conditions; the preset speed increasing step length can be understood as a ratio of a rotating speed signal duty ratio of the fan module every time, and can be determined according to practical application requirements, for example, if the preset speed increasing step length is 10%, after detecting that the temperature of the CPU module or the memory module reaches a preset protection temperature threshold point, the fan rotating speed can be increased by 10% to adjust, so as to ensure reliable working environments of the CPU and the memory, if the temperature of the CPU module or the memory module is not reduced to a corresponding safety threshold after the rotating speed of the fan is increased for a period of time, the fan rotating speed is continuously increased by 10%, and a temperature monitoring result after a period of time is waited for processing until the polling temperature of the CPU module is reduced to be within a first preset temperature threshold range, and the polling temperature of the memory module is also reduced to be within a second preset temperature threshold range; after determining that the temperatures of the CPU module and the memory module are both reduced to be within the corresponding safe ranges, in order to avoid overload of the power PSU, the rotational speed of each fan in the fan module is reduced to the third preset value again, that is, the power consumption of the fan module is reduced again;
It should be noted that, the above process of dynamically adjusting the running power consumption of the fan module according to the CPU polling temperature and the memory polling temperature may be understood as an adjusting process that ensures dynamic balance between the system power consumption and the module temperature at different times as much as possible, so as to effectively ensure the normal operation of the server system in the low power consumption mode, and further provide reliable guarantee for the persistence of the system service processing.
According to the embodiment, after the triggering power consumption reduction event is monitored, the power consumption of the whole machine can be rapidly reduced within ten milliseconds, the system downtime caused by PSU overload is avoided, and the stability of a server system and the continuity of service are maintained; in order to further prove the reliability and the high efficiency of the overall power consumption reduction method provided by the embodiment, the embodiment also uses the BMC to monitor the power failure event as an example for verification and explanation, and obtains the power consumption change statistical result shown in table 2 and the power consumption change trend chart shown in fig. 5:
table 2 Power consumption variation statistics for executing the complete machine Power consumption reduction method of the present invention when BMC monitors Power failure
As shown in fig. 5 and table 2: at 0ms (millisecond) the BMC monitors for power failure (PSU Fail); in the period of 3 ms-5 ms, the BMC rapidly reduces the power consumption of the CPU and the fan by controlling the rotating speed of the fan and controlling the frequency-reducing of the CPU, the total power consumption of the system is 1360 watts and is about 66% of the peak value, and the residual PSU can completely bear the total power consumption of the system; with the reduction of the CPU performance, the load of operation and access to equipment such as a memory, a hard disk, a PCIe card and the like is reduced, the power consumption of the equipment such as the memory, the hard disk, the PCIe card and the like is gradually reduced, and finally the system power consumption is adjusted at 13ms, so that the system is kept in a relatively constant working state, and the normal operation of the system is ensured not to be down. In the adjustment process, the rotation speed of the fan is adjusted to be the lowest at the beginning 0-3 ms, the power consumption of other components is not reduced to the lowest, the temperature of the system is increased to some extent, the rotation speed of the fan is adjusted to some extent, and the rotation speed of the fan is increased to some extent at the 9-13 ms stage, but the overall power consumption is effectively reduced.
The embodiment can effectively verify the simplicity, high efficiency and reliability of the provided complete machine power consumption reduction scheme, and in principle, the requirement of a server system for quickly reducing power consumption in a special operation scene in practical application can be met through the method, but considering that the server system needs to be restored to a normal working mode in time after the power PSU fault is relieved and the operation and maintenance monitoring requirement of a network terminal is finished, the high-efficiency processing capability of the server system can be ensured, and timeliness and high efficiency of service processing are further ensured; specifically, the step of dynamically adjusting the system power consumption according to the power supply capability of the power supply further includes:
When the system running state is in a low-power-consumption working mode, monitoring and recovering normal power consumption events in real time through the baseboard management controller; the normal power consumption recovery event corresponds to the triggering power consumption reduction event, and comprises normal power supply of a power supply or a request for the normal power consumption recovery sent by a network terminal, wherein the normal power supply of the power supply can be understood as power failure recovery, and the power supply is in place and has normal power supply capacity; the network terminal sending a request for restoring normal power consumption can be understood as a request for quickly restoring normal power consumption, which is sent by relevant operation and maintenance personnel or a machine room control end through a network and is received and processed by the BMC;
When the baseboard management controller monitors the normal power consumption recovery event, the system running state is adjusted to a normal working mode, the CPU module and the fan module are adjusted to corresponding normal running power consumption, and the system temperature polling interval is recovered to a normal temperature polling interval; the adjustment of the CPU module and the fan module to the corresponding normal running power consumption may be understood as that the rotational speed of each fan unit is adjusted to the rotational speed in the normal working mode and is continuously maintained, and at the same time, the corresponding cTDP value of the CPU is adjusted to the normal power consumption value or a frequency-reducing (CPU thread) signal is released, and the specific adjustment method may be implemented by referring to the related description of the foregoing power consumption reduction part, which is not repeated herein; meanwhile, the normal temperature polling interval can be understood as a temperature polling interval of the server system working in a normal working mode, and can be set according to practical application conditions, and is not particularly limited herein;
According to the embodiment of the application, the system operation state is set to be the normal operation mode through responding to the starting operation of the server system, the components of the system are set to be the corresponding normal operation power consumption through the baseboard management controller, the trigger power consumption reduction event is monitored in real time, the system operation state is adjusted to be the low power consumption operation mode when the baseboard management controller monitors the trigger power consumption reduction event, after the preset time, the system power consumption is dynamically adjusted according to the power supply capacity until the system is shut down, the technical scheme that the baseboard management controller is combined with the preset trigger power consumption reduction event and the normal power consumption restoration event to maintain and update the system operation state is realized, the dynamic adjustment of the CPU module and the fan module according to the power supply capacity is realized, compared with the prior art, the system operation state is effectively applied to the power failure scene, the system downtime caused by PSU is avoided, the continuity of the system is effectively maintained, the power consumption dynamic adjustment requirement of the network terminal is met, the system operation state is convenient for more operation maintenance and maintenance of the system is also ensured to be high, and the service value is further ensured to be high after the power supply capacity is applied to the server system operation state is timely.
Although the steps in the flowcharts described above are shown in order as indicated by arrows, these steps are not necessarily executed in order as indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders.
In one embodiment, as shown in fig. 6, there is provided a system for reducing power consumption of a whole machine, the system including:
The initialization module 1 is used for responding to the starting operation of the server system, setting the system operation state into a normal operation mode through the baseboard management controller, setting each part of the system into corresponding normal operation power consumption, and monitoring and triggering a power consumption reduction event in real time;
And the power consumption adjusting module 2 is used for adjusting the running state of the system to a low-power consumption working mode when the baseboard management controller monitors the trigger power consumption reducing event, and dynamically adjusting the power consumption of the system according to the power supply capacity after a preset time length until the system is shut down.
The specific limitation of a system for reducing power consumption of a complete machine may be referred to above as limitation of a method for reducing power consumption of a complete machine, and corresponding technical effects may be equally obtained, which is not described herein. The above-mentioned various modules in the whole machine power consumption reduction system can be implemented in whole or in part by software, hardware and combinations thereof. The above modules may be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor may call and execute operations corresponding to the above modules.
In one embodiment, a baseboard management controller is provided that includes a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the method described above when executing the computer program.
In one embodiment, a server is provided, comprising: the baseboard management controller provided by the embodiment is connected with the CPU module, the fan module, the memory module, the hard disk module and the PCIe card.
In summary, the method for reducing the overall power consumption, the system, the baseboard management controller and the server provided by the embodiment of the invention realize the starting-up operation of the server system in response to the overall power consumption reduction method, the baseboard management controller sets the system operation state to be a normal operation mode, sets each part of the system to be corresponding normal operation power consumption, monitors the triggering power consumption reduction event in real time, adjusts the system operation state to be a low power consumption operation mode when the baseboard management controller monitors the triggering power consumption reduction event, and dynamically adjusts the system power consumption according to the power supply capacity after a preset time period until the system is shut down.
In this specification, each embodiment is described in a progressive manner, and all the embodiments are directly the same or similar parts referring to each other, and each embodiment mainly describes differences from other embodiments. In particular, for system embodiments, since they are substantially similar to method embodiments, the description is relatively simple, as relevant to see a section of the description of method embodiments. It should be noted that, any combination of the technical features of the foregoing embodiments may be used, and for brevity, all of the possible combinations of the technical features of the foregoing embodiments are not described, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The foregoing examples represent only a few preferred embodiments of the present application, which are described in more detail and are not to be construed as limiting the scope of the application. It should be noted that modifications and substitutions can be made by those skilled in the art without departing from the technical principles of the present application, and such modifications and substitutions should also be considered to be within the scope of the present application. Therefore, the protection scope of the patent of the application is subject to the protection scope of the claims.

Claims (8)

1. The method for reducing the power consumption of the whole machine is characterized by comprising the following steps of:
Responding to the starting operation of the server system, setting the operation state of the system into a normal operation mode through a baseboard management controller, setting each component of the system into corresponding normal operation power consumption, and monitoring and triggering a power consumption reduction event in real time; the triggering power consumption reduction event comprises power failure, power supply failure or a power consumption reduction request sent by a network terminal;
When the baseboard management controller monitors the trigger power consumption reduction event, the system running state is adjusted to a low power consumption working mode, and after a preset time length, the system power consumption is dynamically adjusted according to the power supply capacity of the power supply until the system receives a shutdown command to shutdown;
The step of dynamically adjusting the system power consumption according to the power supply capacity of the power supply comprises the following steps:
According to the power supply capacity of the power supply, carrying out power consumption reduction adjustment on the pre-screened power consumption module to be reduced; the power consumption reduction adjustment comprises power consumption adjustment based on dynamic balance between system power consumption and module temperature;
the pre-screened power consumption module to be reduced comprises a CPU module and a fan module; the screening process of the power consumption module to be reduced comprises the following steps: and respectively acquiring the power consumption information of each functional module of the server under the condition of no load and load by the baseboard management controller, and screening to obtain the power consumption module to be reduced according to the summarized statistical result of the power consumption information and the actual working principle of the server.
2. The method for reducing power consumption of a complete machine according to claim 1, wherein the step of performing power consumption reduction adjustment on the pre-screened power consumption module to be reduced according to the power supply capability of the power supply comprises:
And transmitting a frequency-reducing command to a CPU module interface through the baseboard management controller, and reducing the duty ratio of a rotating speed signal corresponding to the fan module.
3. The overall power consumption reduction method as claimed in claim 2, wherein the step of transmitting the down-conversion command to the CPU module interface through the baseboard management controller includes:
Reducing cTDP register values corresponding to the CPU module to a first preset value; or, pulling down the CPU down-conversion signal to a second preset value through the general input/output port;
acquiring power output power consumption and current system power consumption in real time, and judging whether the power output power consumption is larger than the current system power consumption or not;
If yes, the cTDP register value corresponding to the CPU module is increased to the maximum allowable CPU frequency value, otherwise, the cTDP register value corresponding to the CPU module is kept unchanged.
4. The method for reducing power consumption of a complete machine as set forth in claim 3, wherein the step of reducing, by the baseboard management controller, a duty cycle of a rotational speed signal corresponding to the fan module comprises:
Reducing pulse width modulation registers of all fan rotating speed controllers corresponding to the fan modules to a third preset value;
shortening the system temperature polling interval, and acquiring the CPU polling temperature and the memory polling temperature of the system through the baseboard management controller;
Judging whether a preset working condition is met or not according to the CPU polling temperature and the memory polling temperature; the preset working condition is that the CPU polling temperature is within a first preset temperature threshold value, and the memory polling temperature is within a second preset temperature threshold value;
If not, the duty ratio of the rotating speed signal corresponding to the fan module is gradually increased according to a preset speed increasing step length until the CPU polling temperature and the memory polling temperature meet the preset working conditions, and after the CPU polling temperature and the memory polling temperature meet the preset working conditions, the pulse width modulation register of each fan rotating speed controller corresponding to the fan module is reduced to a third preset value.
5. The method for reducing power consumption of a complete machine according to claim 4, wherein said step of dynamically adjusting system power consumption according to power supply capacity further comprises:
When the system running state is in a low-power-consumption working mode, monitoring and recovering normal power consumption events in real time through the baseboard management controller;
When the baseboard management controller monitors the normal power consumption recovery event, the system running state is adjusted to a normal working mode, the CPU module and the fan module are adjusted to corresponding normal running power consumption, and the system temperature polling interval is recovered to a normal temperature polling interval.
6. A system for reducing overall power consumption, the system comprising:
The initialization module is used for responding to the starting operation of the server system, setting the system operation state into a normal operation mode, setting each component of the system into corresponding normal operation power consumption, and monitoring and triggering a power consumption reduction event in real time; the triggering power consumption reduction event comprises power failure, power supply failure or a power consumption reduction request sent by a network terminal;
The power consumption adjusting module is used for adjusting the running state of the system into a low-power consumption working mode when the trigger power consumption reducing event is monitored, and dynamically adjusting the power consumption of the system according to the power supply capacity of the power supply after a preset time length until the system is shut down;
wherein, the dynamic adjustment of the system power consumption according to the power supply capability of the power supply comprises:
According to the power supply capacity of the power supply, carrying out power consumption reduction adjustment on the pre-screened power consumption module to be reduced; the power consumption reduction adjustment comprises power consumption adjustment based on dynamic balance between system power consumption and module temperature;
The pre-screened power consumption module to be reduced comprises a CPU module and a fan module; the screening process of the power consumption module to be reduced comprises the following steps: and respectively acquiring the power consumption information of each functional module of the server under the condition of no load and load by using the substrate management controller, and screening to obtain the power consumption module to be reduced according to the summarized statistical result of the power consumption information and the actual working principle of the server.
7. A baseboard management controller comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method according to any one of claims 1 to 5 when executing the computer program.
8. A server, comprising:
The baseboard management controller of claim 7, wherein the CPU module, the fan module, the memory module, the hard disk module, and the PCIe card are connected to the baseboard management controller.
CN202310715603.9A 2023-06-16 2023-06-16 Whole machine power consumption reduction method, system, substrate management controller and server Active CN116755542B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310715603.9A CN116755542B (en) 2023-06-16 2023-06-16 Whole machine power consumption reduction method, system, substrate management controller and server

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310715603.9A CN116755542B (en) 2023-06-16 2023-06-16 Whole machine power consumption reduction method, system, substrate management controller and server

Publications (2)

Publication Number Publication Date
CN116755542A CN116755542A (en) 2023-09-15
CN116755542B true CN116755542B (en) 2024-06-04

Family

ID=87949124

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310715603.9A Active CN116755542B (en) 2023-06-16 2023-06-16 Whole machine power consumption reduction method, system, substrate management controller and server

Country Status (1)

Country Link
CN (1) CN116755542B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117590923B (en) * 2024-01-19 2024-06-18 北京卡普拉科技有限公司 Method, device and equipment for reducing power consumption based on I/O access

Citations (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103246590A (en) * 2013-04-22 2013-08-14 山东超越数控电子有限公司 Multiunit server fan monitoring system
CN106292963A (en) * 2016-08-24 2017-01-04 浪潮电子信息产业股份有限公司 Fan regulation and control method based on CPU power consumption
CN106640716A (en) * 2015-07-31 2017-05-10 中兴通讯股份有限公司 Fan control method and device capable of giving consideration to all service panels
CN110147155A (en) * 2019-05-21 2019-08-20 苏州浪潮智能科技有限公司 Server power supply cold redundancy control method, device and BMC based on BMC
CN110362175A (en) * 2019-06-29 2019-10-22 苏州浪潮智能科技有限公司 A kind of control method for fan and device
CN110836197A (en) * 2019-11-05 2020-02-25 英业达科技有限公司 Strategy for automatically optimizing energy consumption
CN111475288A (en) * 2020-03-27 2020-07-31 苏州浪潮智能科技有限公司 Server and power supply protection system thereof
CN113835508A (en) * 2021-09-27 2021-12-24 苏州浪潮智能科技有限公司 Power consumption management method and device of server, BMC and storage medium
CN114442787A (en) * 2022-01-28 2022-05-06 苏州浪潮智能科技有限公司 Method and system for realizing whole machine power consumption callback after server enters power consumption capping
CN114721488A (en) * 2022-03-18 2022-07-08 上海燧原科技有限公司 Fan rotating speed control method and system
CN115686172A (en) * 2022-10-25 2023-02-03 苏州浪潮智能科技有限公司 Power consumption control method and device, electronic equipment, readable storage medium and product
WO2023029375A1 (en) * 2021-09-03 2023-03-09 苏州浪潮智能科技有限公司 Power source consumption management apparatus for four-way server
CN115904050A (en) * 2022-12-15 2023-04-04 苏州浪潮智能科技有限公司 Power control system and method for preventing power failure shutdown of server
CN115904055A (en) * 2022-11-08 2023-04-04 苏州浪潮智能科技有限公司 Power saving method, device, equipment and storage medium based on baseboard management controller

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8756440B2 (en) * 2008-04-16 2014-06-17 International Business Machines Corporation System power capping using information received from the installed power supply
US10459733B2 (en) * 2018-01-25 2019-10-29 Dell Products, Lp Power budgeting in an information handling system
US10877553B2 (en) * 2018-04-13 2020-12-29 Dell Products L.P. Systems and methods for power control based on power consumption of storage devices
TWI675288B (en) * 2018-09-21 2019-10-21 神雲科技股份有限公司 Server rack
CN109236714B (en) * 2018-11-06 2020-08-07 英业达科技有限公司 Fan control circuit and fan control system
US11409560B2 (en) * 2019-03-28 2022-08-09 Intel Corporation System, apparatus and method for power license control of a processor

Patent Citations (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103246590A (en) * 2013-04-22 2013-08-14 山东超越数控电子有限公司 Multiunit server fan monitoring system
CN106640716A (en) * 2015-07-31 2017-05-10 中兴通讯股份有限公司 Fan control method and device capable of giving consideration to all service panels
CN106292963A (en) * 2016-08-24 2017-01-04 浪潮电子信息产业股份有限公司 Fan regulation and control method based on CPU power consumption
CN110147155A (en) * 2019-05-21 2019-08-20 苏州浪潮智能科技有限公司 Server power supply cold redundancy control method, device and BMC based on BMC
CN110362175A (en) * 2019-06-29 2019-10-22 苏州浪潮智能科技有限公司 A kind of control method for fan and device
CN110836197A (en) * 2019-11-05 2020-02-25 英业达科技有限公司 Strategy for automatically optimizing energy consumption
CN111475288A (en) * 2020-03-27 2020-07-31 苏州浪潮智能科技有限公司 Server and power supply protection system thereof
WO2023029375A1 (en) * 2021-09-03 2023-03-09 苏州浪潮智能科技有限公司 Power source consumption management apparatus for four-way server
CN113835508A (en) * 2021-09-27 2021-12-24 苏州浪潮智能科技有限公司 Power consumption management method and device of server, BMC and storage medium
CN114442787A (en) * 2022-01-28 2022-05-06 苏州浪潮智能科技有限公司 Method and system for realizing whole machine power consumption callback after server enters power consumption capping
CN114721488A (en) * 2022-03-18 2022-07-08 上海燧原科技有限公司 Fan rotating speed control method and system
CN115686172A (en) * 2022-10-25 2023-02-03 苏州浪潮智能科技有限公司 Power consumption control method and device, electronic equipment, readable storage medium and product
CN115904055A (en) * 2022-11-08 2023-04-04 苏州浪潮智能科技有限公司 Power saving method, device, equipment and storage medium based on baseboard management controller
CN115904050A (en) * 2022-12-15 2023-04-04 苏州浪潮智能科技有限公司 Power control system and method for preventing power failure shutdown of server

Also Published As

Publication number Publication date
CN116755542A (en) 2023-09-15

Similar Documents

Publication Publication Date Title
CN116755542B (en) Whole machine power consumption reduction method, system, substrate management controller and server
US20030023885A1 (en) Automated power management system for a network of computers
TW201329690A (en) Server rack system and power management method applicable thereto
CN111475288A (en) Server and power supply protection system thereof
CN106371540B (en) System power management method, chip and electronic equipment
CN109236710B (en) Server fan control system and control method thereof
CN103699471A (en) Method and device for controlling temperature of computer hardware
CN110442225B (en) Power distribution board, modular chassis system and operation method thereof
CN115344453A (en) Server running state monitoring device, method, computer equipment and storage medium
US20100318826A1 (en) Changing Power States Of Data-Handling Devices To Meet Redundancy Criterion
WO2023029375A1 (en) Power source consumption management apparatus for four-way server
CN111309132B (en) Method for multi-gear power supply redundancy of server
CN117331425B (en) Power consumption management system, power consumption management method, storage medium, and electronic device
CN111984471B (en) Cabinet power BMC redundancy management system and method
CN116823587A (en) Graphics processor control method and device, electronic equipment and storage medium
CN117453036A (en) Method, system and device for adjusting power consumption of equipment in server
CN115904050A (en) Power control system and method for preventing power failure shutdown of server
CN116991221A (en) Power consumption adjusting method and device
US11093014B2 (en) Method for monitoring, control and graceful shutdown of control and/or computer units
CN111026252A (en) Method and device for server temperature redundancy control
CN111045844A (en) Fault degradation method and device
CN114610562B (en) Server power consumption management and control method, system, terminal and storage medium
CN112394803B (en) Domestic server power consumption real-time dynamic adjusting method
CN111475293B (en) Server and power supply protection system thereof
CN111752366B (en) PSU abnormal power failure processing system, method and device

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