Intelligent energy-saving case of data center

Abstract

According to the data center intelligent energy-saving case, the historical energy consumption data and the current energy consumption data of the data center are determined, the current energy consumption data and the historical energy consumption data are determined, the power consumption state is determined according to the current energy consumption data, and when the representation of the power consumption state is unreasonable, the power consumption state of the data center is determined. According to the current energy consumption data and the historical energy consumption data, determining predicted server energy consumption, predicted air conditioner energy consumption and predicted storage equipment energy consumption, providing a prediction mechanism to predict energy consumption in advance, and according to the predicted server energy consumption, the predicted air conditioner energy consumption and the predicted storage equipment energy consumption, providing a prediction mechanism to predict energy consumption in advance; at least one of server parameters, air conditioner parameters and storage equipment parameters is adjusted according to the power consumption state, so that the power consumption state characterization is reasonable, and in-advance adjustment and prevention, energy consumption reduction and power saving are facilitated when an unreasonable power consumption state with excessive consumption possibly occurs.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of data centers, in particular to a data center intelligent energy-saving case. BACKGROUND

[0002] With the development of artificial intelligence and cloud computing technology, the requirements for data centers for providing computing power support are also increasing, one aspect of which is the energy-saving demand for data centers. At present, for the energy consumption control of data centers, it is usually based on real-time data collection to calculate energy consumption or through manual inspection to determine whether there is an energy consumption anomaly, and take corresponding measures to achieve the purpose of energy saving. On the one hand, it cannot prevent in advance, and the timeliness is poor, and on the other hand, the efficiency of manual inspection is low. SUMMARY

[0003] The embodiments of the present application provide a data center intelligent energy-saving case to solve at least one problem in the related art, and the technical solutions are as follows:

[0004] In a first aspect, the embodiments of the present application provide a data center intelligent energy-saving case, comprising:

[0005] A first determining module is configured to determine historical energy consumption data and current energy consumption data of the data center, wherein the current energy consumption data and the historical energy consumption data each include server energy consumption, air conditioner energy consumption, and storage device energy consumption;

[0006] A second determining module is configured to determine a power consumption state according to the current energy consumption data;

[0007] A prediction module is configured to, when the power consumption state is unreasonable, determine predicted server energy consumption, predicted air conditioner energy consumption, and predicted storage device energy consumption according to the current energy consumption data and the historical energy consumption data;

[0008] An adjusting module is configured to, according to the predicted server energy consumption, the predicted air conditioner energy consumption, and the predicted storage device energy consumption, adjust at least one of server parameters, air conditioner parameters, and storage device parameters, so that the power consumption state is reasonable.

[0009] In an embodiment, the determination of the current energy consumption data of the data center comprises:

[0010] The chip idle power consumption, the chip full-load power consumption, and the chip utilization rate of at least one chip are obtained, the ratio of the chip utilization rate to a preset value is calculated, the adjustment term is determined according to the ratio and a nonlinear coefficient, the first product of the power difference between the chip full-load power consumption and the chip idle power consumption and the adjustment term is determined, and the current chip power consumption is obtained according to the sum of the first product and the chip idle power consumption. The server energy consumption is determined according to the sum of the current chip power consumptions of all chips.

[0011] The air conditioner energy consumption is obtained through the ammeter device;

[0012] The active hard disk quantity, the single hard disk active power consumption, the idle hard disk quantity and the single hard disk idle power consumption are obtained, the second product of the active hard disk quantity and the single hard disk active power consumption and the third product of the idle hard disk quantity and the single hard disk idle power consumption are determined, and the storage device energy consumption is determined according to the sum of the second product and the third product.

[0013] In an implementation, the first determining module is further configured to determine the nonlinear coefficient by the following steps:

[0014] The test utilization rate and the test power consumption corresponding to different loads at a specified chip frequency are tested;

[0015] The test utilization rate, the test power consumption, the chip idle power consumption and the chip full load power consumption corresponding to different loads are fitted by using the least square method to determine the basic value of the nonlinear coefficient;

[0016] The type of the task currently performed is determined, and the final nonlinear coefficient is determined according to the type of the task and the basic value of the nonlinear coefficient.

[0017] In an implementation, the final nonlinear coefficient is determined according to the type of the task and the basic value of the nonlinear coefficient, including:

[0018] When the type of the task is calculation training, the basic value is increased by a specified value to determine the final nonlinear coefficient;

[0019] When the type of the task is query, the basic value is decreased by a specified value to determine the final nonlinear coefficient;

[0020] When the type of the task is file reading and writing, transmission or the type of the task cannot be recognized, the size of the basic value is maintained to determine the final nonlinear coefficient.

[0021] In an implementation, the second determining module includes a first unit, a second unit and a third unit:

[0022] The first unit is configured to determine the total power consumption according to the sum of the server energy consumption, the air conditioner energy consumption and the storage device energy consumption, and determine the electronic device energy consumption according to the sum of the server energy consumption and the storage device energy consumption;

[0023] The second unit is configured to determine the power usage efficiency according to the ratio of the electronic device energy consumption to the total power consumption;

[0024] The third unit is configured to determine a state representing reasonable power consumption when the power usage efficiency is greater than or equal to an efficiency threshold, and determine a state representing unreasonable power consumption when the power usage efficiency is less than the efficiency threshold.

[0025] In an implementation, when the power consumption state representation is unreasonable, the predicted server power consumption, the predicted air conditioner power consumption, and the predicted storage device power consumption are determined according to the current energy consumption data and the historical energy consumption data, which comprises:

[0026] When the power consumption state representation is unreasonable, the predicted server power consumption, the predicted air conditioner power consumption, and the predicted storage device power consumption are determined by analyzing the current energy consumption data and the historical energy consumption data through a long short-term memory network.

[0027] The current energy consumption data and the historical energy consumption data both comprise determined time corresponding to the server power consumption, the air conditioner power consumption, and the storage device power consumption.

[0028] In an implementation, the adjustment module comprises a first adjustment unit, a second adjustment unit, a third adjustment unit, a fourth adjustment unit, a fifth adjustment unit, and a sixth adjustment unit.

[0029] The first adjustment unit is configured to determine the predicted total power consumption according to the sum of the predicted server power consumption, the predicted air conditioner power consumption, and the predicted storage device power consumption, and determine the predicted electronic device power consumption according to the sum of the predicted server power consumption and the predicted storage device power consumption.

[0030] The second adjustment unit is configured to determine the predicted power usage efficiency according to the ratio of the predicted electronic device power consumption to the predicted total power consumption, and determine the server proportion, the air conditioner power consumption proportion, and the storage device proportion when the predicted power usage efficiency is less than the efficiency threshold.

[0031] The third adjustment unit is configured to analyze the historical energy consumption data to determine the first average proportion of the server power consumption to the total power consumption, the second average proportion of the air conditioner power consumption to the total power consumption, and the third average proportion of the storage device power consumption to the total power consumption.

[0032] The fourth adjustment unit is configured to reduce the predicted server power consumption when the server proportion is greater than the first average proportion, so that the power consumption state representation is reasonable.

[0033] The fifth adjustment unit is configured to reduce the predicted air conditioner power consumption when the air conditioner power consumption proportion is greater than the second average proportion, so that the power consumption state representation is reasonable.

[0034] The sixth adjustment unit is configured to reduce the predicted storage device power consumption when the storage device proportion is greater than the third average proportion, so that the power consumption state representation is reasonable.

[0035] In an implementation, the reduction of the predicted server power consumption, the reduction of the predicted air conditioner power consumption, and the reduction of the predicted storage device power consumption are specifically:

[0036] at least one of reducing frequency and voltage of the chip, and setting a state of an idle chip to a sleep state;

[0037] at least one of increasing air supply temperature and closing air valves of non-critical areas;

[0038] at least one of reducing RAID level and reducing OP proportion.

[0039] In an embodiment, the adjusting module comprises a seventh adjusting unit, the seventh adjusting unit is configured to:

[0040] when the predicted power usage efficiency is less than the efficiency threshold, starting the backup power supply to supply power cooperatively with the main power supply;

[0041] or,

[0042] when the main power supply state is abnormal, starting the backup power supply to supply power.

[0043] In an embodiment, another data center intelligent energy-saving case is also provided, comprising a processor and a memory, the memory stores instructions, the instructions are loaded and executed by the processor to implement the following method:

[0044] determining historical energy consumption data and current energy consumption data of the data center, the current energy consumption data and the historical energy consumption data both comprise server energy consumption, air conditioner energy consumption and storage device energy consumption;

[0045] determining a power consumption state according to the current energy consumption data;

[0046] when the power consumption state representation is unreasonable, determining predicted server energy consumption, predicted air conditioner energy consumption and predicted storage device energy consumption according to the current energy consumption data and the historical energy consumption data;

[0047] adjusting at least one of server parameters, air conditioner parameters and storage device parameters according to the predicted server energy consumption, the predicted air conditioner energy consumption and the predicted storage device energy consumption, so that the power consumption state representation is reasonable.

[0048] The above technical solutions have at least the following beneficial effects:

[0049] By determining the historical energy consumption data and the current energy consumption data of the data center, the current energy consumption data and the historical energy consumption data, determining the power consumption state according to the current energy consumption data, when the power consumption state represents unreasonable, determining the predicted server energy consumption, the predicted air conditioner energy consumption and the predicted storage device energy consumption according to the current energy consumption data and the historical energy consumption data, providing a prediction mechanism to predict energy consumption in advance, adjusting at least one of the server parameters, the air conditioner parameters and the storage device parameters according to the predicted server energy consumption, the predicted air conditioner energy consumption and the predicted storage device energy consumption, so that the power consumption state represents reasonable, which is beneficial to adjust in advance and prevent when the unreasonable power consumption state of excessive consumption may occur, reduce energy consumption and save power.

[0050] The above summary is merely intended to illustrate the present description and is not intended to limit in any way. In addition to the illustrative aspects, embodiments and features described above, further aspects, embodiments and features will be readily apparent to those skilled in the art by reference to the drawings and the following detailed description. BRIEF DESCRIPTION OF DRAWINGS

[0051] In the drawings, like numerals refer to like elements throughout the various drawings. The drawings are not necessarily to scale, the emphasis instead being placed on illustrating principles of the present application. It should be understood that the drawings are merely depictions of some embodiments of the present application and should not be construed as limiting the present application.

[0052] Figure 1 The structural block diagram of the data center intelligent energy-saving machine case according to an embodiment of the present application is shown in the figure.

[0053] Figure 2 The structural block diagram of the data center intelligent energy-saving machine case according to another embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0054] In the following, only some exemplary embodiments are simply described. As those skilled in the art can recognize, the described embodiments can be modified in various different ways without departing from the spirit or scope of the present application. Therefore, the drawings and the description are considered to be exemplary in nature rather than limiting.

[0055] Reference Figure 1 The structural block diagram of the data center intelligent energy-saving machine case according to an embodiment of the present application is shown in the figure, which has a sensor module, a processor chip and a display screen; the processor chip as the main processing core includes:

[0056] The first determining module is configured to determine the historical energy consumption data and the current energy consumption data of the data center, and the current energy consumption data and the historical energy consumption data both include the server energy consumption, the air conditioner energy consumption and the storage device energy consumption;

[0057] a second determining module configured to determine a power consumption state according to the current energy consumption data;

[0058] a predicting module configured to, when the power consumption state is unreasonable, determine predicted server energy consumption, predicted air conditioner energy consumption and predicted storage device energy consumption according to the current energy consumption data and the historical energy consumption data;

[0059] an adjusting module configured to, according to the predicted server energy consumption, the predicted air conditioner energy consumption and the predicted storage device energy consumption, adjust at least one of server parameters, air conditioner parameters and storage device parameters, so as to make the power consumption state reasonable.

[0060] The technical scheme of the embodiment of the present application determines the historical energy consumption data and the current energy consumption data of the data center, determines the power consumption state according to the current energy consumption data when the power consumption state is unreasonable, determines the predicted server energy consumption, the predicted air conditioner energy consumption and the predicted storage device energy consumption according to the current energy consumption data and the historical energy consumption data, provides a prediction mechanism to predict energy consumption in advance, adjusts at least one of the server parameters, the air conditioner parameters and the storage device parameters according to the predicted server energy consumption, the predicted air conditioner energy consumption and the predicted storage device energy consumption, so as to make the power consumption state reasonable, which is beneficial to adjusting and preventing in advance when the unreasonable power consumption state of excessive consumption occurs, reduces energy consumption and saves power.

[0061] In an implementation manner, the first determining module can obtain the historical energy consumption data of the data center from the operation log of the data center, and obtain the current energy consumption data of the data center by monitoring the monitoring data of the data center in real time. Optionally, the data types of the current energy consumption data and the historical energy consumption data are similar, and both of them include server energy consumption, air conditioner energy consumption and storage device energy consumption. Other energy consumptions can be considered in other implementation manners, which are not limited specifically.

[0062] In an implementation manner, the monitoring data includes but is not limited to at least one of the following: chip idle power consumption p idle of a chip (for example, CPU, GPU, etc., the CPU is taken as an example in the embodiment of the present application), chip full load power consumption p max , chip utilization rate u cpu , number of active hard disks N active , single hard disk active power consumption P active , number of idle hard disks N idle and single hard disk idle power consumption P idle , etc., which are not limited specifically.

[0063] Optionally, the current energy consumption data of the data center is determined, including S101-S103:

[0064] S101, obtain the chip idle power consumption, the chip full load power consumption, and the chip utilization rate of at least one chip, calculate the ratio of the chip utilization rate to the preset value, determine the adjustment term according to the ratio and the nonlinear coefficient, determine the first product of the power difference between the chip full load power consumption and the chip idle power consumption and the adjustment term, and obtain the current chip power consumption according to the sum of the first product and the chip idle power consumption, determine the server energy consumption according to the sum of the current chip power consumptions of all chips.

[0065] For example, the preset value is 100, and the nonlinear coefficient k is taken as an example to determine the ratio of the chip utilization rate to the preset value According to the ratio and the nonlinear coefficient k, the adjustment term is determined Then, the power difference (p max -p idle ) between the chip full load power consumption and the chip idle power consumption is determined And the first product of the sum of the first product and the chip idle power consumption p idle , the current chip power consumption p 芯片 is obtained, and the server energy consumption p 服务器 is determined according to the sum of the current chip power consumptions of all chips.

[0066]

[0067] In an embodiment, the nonlinear coefficient k is determined by the following steps S110-S130:

[0068] S110, test the corresponding test utilization rate and test power consumption under different loads at a specified chip frequency.

[0069] Optionally, dynamic frequency adjustment is disabled and a specified chip frequency is used, which can be a certain proportion of the maximum chip frequency, based on actual needs adjustment, and then the corresponding test utilization rate u' cpu and test power consumption p' under different loads are tested by the existing method, for example, gradually increasing the load from 1%, 10%, 20%, …, 100%.

[0070] S120, the corresponding test utilization rate u' cpu , test power consumption p', chip idle power consumption p idle , and chip full load power consumption p max under different loads are fitted by least squares method to determine the basic value of the nonlinear coefficient.

[0071] Optionally, the least squares fitting formula is:

[0072]

[0073] The final x value obtained by the least squares fitting is used as the basic value of the nonlinear coefficient k, that is, the nonlinear coefficient k=x.

[0074] S130: Determine the type of the currently performed task, and determine the final nonlinear coefficient according to the task type and the basic value of the nonlinear coefficient.

[0075] Optionally, in order to improve the accuracy of the nonlinear coefficient k, targeted personalized adjustments are made based on the task type. Specifically, based on the task type and the basic value of the nonlinear coefficient, the final nonlinear coefficient, i.e., the final value of the nonlinear coefficient k, is determined, including:

[0076] 1. When the task type is calculation training, increase the base value by the specified value to determine the final nonlinear coefficient.

[0077] It should be noted that computational training includes but is not limited to tasks such as matrix calculations and model training, which require high chip utilization; queries include but are not limited to database queries, which generally require low chip utilization. Among them, tasks may also involve task types such as file reading and writing, transmission, or other task types that are not pre-set and cannot be identified. Optionally, when the task type is computational training, the base value is increased by a specified value to determine the final nonlinear coefficient k; wherein the size of the specified value is set based on the actual situation.

[0078] 2. When the task type is query, reduce the base value by the specified value to determine the final nonlinear coefficient k.

[0079] Optionally, when the task type is query, the base value is reduced by a specified value to determine the final nonlinear coefficient.

[0080] 3. When the task type is file reading, writing, or transmission, or the task type cannot be identified, maintain the size of the base value and determine the final nonlinear coefficient.

[0081] Optionally, since the chip utilization requirements for file reading, writing, and transmission are medium and the task type cannot be identified, a balanced approach is adopted, thereby maintaining the size of the base value and determining the final nonlinear coefficient k.

[0082] S102. Obtain air conditioning energy consumption through an electric meter.

[0083] Optionally, the sensor module may include an electric meter device, which obtains the air conditioner energy consumption p 空调 Then it is transmitted to the processor chip for processing.

[0084] S103, obtaining the number of active hard disks, the active power consumption of a single hard disk, the number of idle hard disks and the idle power consumption of a single hard disk, determining a second product of the number of active hard disks and the active power consumption of a single hard disk and a third product of the number of idle hard disks and the idle power consumption of a single hard disk, and determining the energy consumption of the storage device according to the sum of the second product and the third product.

[0085] Optionally, the number of active hard disks N active , the active power consumption of a single hard disk P active , the number of idle hard disks N idle , and the idle power consumption of a single hard disk P idle It should be noted that the above data can be obtained by existing methods, which will not be described herein. The formula for calculating the energy consumption of the storage device p 存储设备 is as follows:

[0086] p 存储设备 = P active * N active + P idle * N idle

[0087] In the embodiments of the present application, the second determining module includes a first unit, a second unit and a third unit.

[0088] The first unit is configured to determine the total power consumption according to the sum of the server energy consumption, the air conditioner energy consumption and the storage device energy consumption, and determine the electronic device energy consumption according to the sum of the server energy consumption and the storage device energy consumption.

[0089] The second unit is configured to determine the power usage efficiency according to the ratio of the electronic device energy consumption to the total power consumption.

[0090] The total power consumption p 总 = p 服务器 + p 空调 + p 存储设备 , the electronic device energy consumption p 电子设备 = p 服务器 + p 存储设备 , and the power usage efficiency = p 电子设备 / p 总 .

[0091] The third unit is configured to determine a reasonable power consumption state when the power usage efficiency is greater than or equal to an efficiency threshold, and determine an unreasonable power consumption state when the power usage efficiency is less than the efficiency threshold.

[0092] Optionally, if the power usage efficiency is greater than or equal to the efficiency threshold, a reasonable power consumption state is determined, and if the power usage efficiency is less than the efficiency threshold, the current power consumption state is determined to be an unreasonable power consumption state, which needs to be adjusted.

[0093] In an implementation, when the power consumption state is unreasonable, the prediction module determines the predicted server energy consumption, the predicted air conditioner energy consumption and the predicted storage device energy consumption according to the current energy consumption data and the historical energy consumption data, specifically, by analyzing the current energy consumption data and the historical energy consumption data through a long short-term memory network (LSTM) to determine the predicted server energy consumption, the predicted air conditioner energy consumption and the predicted storage device energy consumption.

[0094] It should be noted that the current energy consumption data and the historical energy consumption data both include the determined time corresponding to the server energy consumption, the air conditioner energy consumption and the storage device energy consumption, which serves as the basis for training and prediction of the long short-term memory network (LSTM).

[0095] In an implementation, the adjustment module includes a first adjustment unit, a second adjustment unit, a third adjustment unit, a fourth adjustment unit, a fifth adjustment unit and a sixth adjustment unit.

[0096] The first adjustment unit is configured to determine the predicted total power consumption according to the sum of the predicted server energy consumption, the predicted air conditioner energy consumption and the predicted storage device energy consumption, and determine the predicted electronic device energy consumption according to the sum of the predicted server energy consumption and the predicted storage device energy consumption.

[0097] The second adjustment unit is configured to determine the predicted power usage efficiency according to the ratio of the predicted electronic device energy consumption to the predicted total power consumption, and when the predicted power usage efficiency is less than an efficiency threshold, determine the proportion of the predicted server energy consumption, the predicted air conditioner energy consumption and the predicted storage device energy consumption in the predicted total power consumption respectively to obtain a server proportion, an air conditioner energy consumption proportion and a storage device proportion.

[0098] Optionally, when the predicted power usage efficiency is less than the efficiency threshold, the ratio of the predicted server energy consumption, the predicted air conditioner energy consumption and the predicted storage device energy consumption to the predicted total power consumption is calculated respectively to obtain the server proportion, the air conditioner energy consumption proportion and the storage device proportion respectively.

[0099] The third adjustment unit is configured to analyze the historical energy consumption data to determine a first average proportion of the server energy consumption in the total power consumption, a second average proportion of the air conditioner energy consumption in the total power consumption and a third average proportion of the storage device energy consumption in the total power consumption respectively.

[0100] Similarly, based on the analysis of historical energy consumption data, the time periods are divided into specified time intervals to determine the first proportion of server energy consumption in the total power consumption, the second proportion of air conditioning energy consumption in the total power consumption, and the third proportion of storage device energy consumption in the total power consumption in each time period. Then, the ratio of the sum of the first proportions to the number of time periods is calculated to obtain the first average proportion of server energy consumption in the total power consumption, the ratio of the sum of the second proportions to the number of time periods is calculated to obtain the second average proportion of air conditioning energy consumption in the total power consumption, and the ratio of the sum of the third proportions to the number of time periods is calculated to obtain the third average proportion of storage device energy consumption in the total power consumption.

[0101] The fourth adjustment unit is configured to reduce the predicted server energy consumption when the server proportion is greater than the first average proportion, so as to make the representation of the power consumption state reasonable.

[0102] Optionally, when the server proportion is greater than the first average proportion, it is necessary to reduce the predicted server energy consumption to make the power consumption status representation reasonable, such as including but not limited to reducing the frequency and voltage of the chip and setting the state of the idle chip to at least one of the sleep state.

[0103] The fifth adjustment unit is configured to reduce the predicted air conditioning energy consumption when the air conditioning energy consumption proportion is greater than the second average proportion, so as to make the representation of the power consumption state reasonable.

[0104] Optionally, when the proportion of air conditioning energy consumption is greater than the second average proportion, it is necessary to reduce the predicted air conditioning energy consumption to make the power consumption status representation reasonable, such as including but not limited to increasing the supply air temperature and closing at least one of the air valves in non-critical areas. The non-critical areas are determined based on actual conditions and are not specifically limited.

[0105] The sixth adjustment unit is configured to reduce the predicted energy consumption of the storage device when the storage device proportion is greater than the third average proportion, so as to make the representation of the power consumption state reasonable.

[0106] Optionally, when the storage device ratio is greater than the third average ratio, the predicted storage device energy consumption needs to be reduced to achieve a reasonable representation of the power consumption state, including but not limited to at least one of reducing the RAID (Redundant Array of Independent Disks) level (e.g., reducing RAID 10 to RAID 5) and reducing the OP (Over Reserved Space) ratio (e.g., reducing 30% to 20%). Reducing the RAID level can reduce power consumption, while reducing the OP can improve energy efficiency.

[0107] In one embodiment, the adjustment module further includes a seventh adjustment unit, which is configured to:

[0108] When the predicted power efficiency is lower than the efficiency threshold, the backup power supply is turned on to supply power in conjunction with the main power supply;

[0109] or,

[0110] When the main power supply state is abnormal, the standby power supply is started to supply power.

[0111] In the embodiment of the application, by setting the standby power supply on the basis of the main power supply, when the predicted power usage efficiency is less than the efficiency threshold or the main power supply state is abnormal, the standby power supply is used to ensure normal power supply of the data center, so that the data center can operate normally.

[0112] In the embodiment of the application, the display screen can also display the specific details of the calculated server energy consumption, air conditioner energy consumption, storage device energy consumption, predicted server energy consumption, predicted air conditioner energy consumption, predicted storage device energy consumption, task type, adjusted server parameters, adjusted air conditioner parameters, and adjusted storage device parameters, and whether the standby power supply is started or not. By analyzing the power consumption state and the predicted energy consumption, the prediction mechanism is provided and adjusted in advance, unnecessary power consumption can be automatically adjusted, and the purpose of energy saving is achieved.

[0113] Referring to Figure 2 , a structural block diagram of a data center intelligent energy-saving case according to an embodiment of the application is shown, which comprises a memory 310 and a processor 320. The memory 310 stores instructions executable on the processor 320. The processor 320 loads and executes the instructions to implement the following method.

[0114] Determine historical energy consumption data and current energy consumption data of the data center. The current energy consumption data and the historical energy consumption data both include server energy consumption, air conditioner energy consumption, and storage device energy consumption.

[0115] Determine the power consumption state according to the current energy consumption data.

[0116] When the power consumption state is unreasonable, determine the predicted server energy consumption, the predicted air conditioner energy consumption, and the predicted storage device energy consumption according to the current energy consumption data and the historical energy consumption data.

[0117] Adjust at least one of the server parameters, the air conditioner parameters, and the storage device parameters according to the predicted server energy consumption, the predicted air conditioner energy consumption, and the predicted storage device energy consumption, so that the power consumption state is reasonable.

[0118] The number of the memory 310 and the processor 320 can be one or more. In an embodiment, the data center intelligent energy-saving case further comprises a communication interface 330 for communicating with external devices, transmitting data, and sharing data. If the memory 310, the processor 320, and the communication interface 330 are independently implemented, the memory 310, the processor 320, and the communication interface 330 can be connected to each other through a bus and complete communication among them.

[0119] In the description of the application, the description of the terms "one embodiment", "some embodiments", "example", "specific example" or "some examples" etc. means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the application. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in one or more embodiments or examples. In addition, different embodiments or examples described in the specification and the features of different embodiments or examples can be combined and combined by those skilled in the art without contradiction, if necessary.

[0120] In addition, the terms "first", "second" are only for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include at least one of the features. In the description of the application, the meaning of "multiple" is two or more, unless otherwise specifically limited.

[0121] Any process or method descriptions in flow charts or described elsewhere herein can be understood as representing code modules, segments, or portions of code that include one or more executable instructions for implementing specific logic functions or other processes. And the various embodiments of the application can include additional or fewer steps or processes in comparison to those shown in the figures.

[0122] The logic and / or steps represented in flow charts or otherwise described herein, for example, can be embodied in computer-readable instructions, which can be used to cause one or more processors to perform the actions indicated in the steps. The computer-readable instructions can be stored on one or more storage media or memory devices associated with the one or more processors.

[0123] It should be understood that parts of the application can be implemented in hardware, software, firmware or a combination thereof. In the above-described embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. All or part of the steps of the above-described embodiment method can be instructed by a program to complete the relevant hardware, which can be stored in a computer-readable storage medium, and the program includes one or a combination of the steps of the method embodiment when executed.

[0124] In addition, each of the function units in each embodiment of the present application can be integrated in one processing module, or each unit can be physically present separately, or two or more units can be integrated in one module. The integrated module can be realized in the form of hardware or in the form of a software function module. When the integrated module is realized in the form of a software function module and sold or used as an independent product, it can also be stored in a computer readable storage medium. The storage medium can be a read-only memory, a magnetic disk or an optical disk, etc.

[0125] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of various changes or replacements within the technical scope disclosed in the present application, and these should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A data center intelligent energy-saving chassis, characterized in that: include: A first determination module is used to determine the historical energy consumption data and current energy consumption data of the data center, where the current energy consumption data and the historical energy consumption data both include server energy consumption, air conditioning energy consumption, and storage device energy consumption; The second determination module determines the power consumption status based on the current energy consumption data; A prediction module is used to determine the predicted server energy consumption, predicted air conditioning energy consumption, and predicted storage device energy consumption based on current energy consumption data and historical energy consumption data when the power consumption status is characterized as unreasonable; The adjustment module is used to adjust at least one of the server parameters, the air conditioning parameters and the storage device parameters according to the predicted server energy consumption, the predicted air conditioning energy consumption and the predicted storage device energy consumption, so as to make the power consumption state representation reasonable.

2. The intelligent energy-saving chassis for a data center according to claim 1, characterized in that: Determining the current energy consumption data of the data center includes: Obtain chip idle power consumption, chip full load power consumption, and chip utilization of at least one chip, calculate the ratio of the chip utilization to a preset value, determine an adjustment item based on the ratio and a nonlinear coefficient, determine a first product of a power difference between the chip full load power consumption and the chip idle power consumption and the adjustment item, obtain the current chip power consumption based on the sum of the first product and the chip idle power consumption, and determine the server energy consumption based on the sum of the current chip power consumptions of all chips; Obtain air conditioning energy consumption through electric meter equipment; Obtain the number of active hard disks, the active power consumption of a single hard disk, the number of idle hard disks, and the idle power consumption of a single hard disk, determine the second product of the number of active hard disks and the active power consumption of a single hard disk, and the third product of the number of idle hard disks and the idle power consumption of a single hard disk, and determine the energy consumption of the storage device based on the sum of the second product and the third product.

3. The intelligent energy-saving chassis for a data center according to claim 2, characterized in that: The first determination module is further configured to determine the nonlinear coefficient by the following steps: At the specified chip frequency, test the corresponding test utilization and test power consumption under different loads; The test utilization, test power consumption, chip idle power consumption, and chip full load power consumption corresponding to different loads are fitted using the least squares method to determine the basic value of the nonlinear coefficient; Determine the type of task currently being performed, and determine the final nonlinear coefficient based on the task type and the basic value of the nonlinear coefficient.

4. The intelligent energy-saving chassis for a data center according to claim 3, characterized in that: Determining the final nonlinear coefficient according to the task type and the basic value of the nonlinear coefficient includes: When the task type is calculation training, the base value is increased by the specified value to determine the final nonlinear coefficient; When the task type is query, the base value is reduced by the specified value to determine the final nonlinear coefficient; When the task type is file reading and writing, transmission, or the task type cannot be identified, the size of the basic value is maintained and the final nonlinear coefficient is determined.

5. The intelligent energy-saving chassis for a data center according to any one of claims 1 to 4, characterized in that: The second determining module includes a first unit, a second unit and a third unit: The first unit is configured to determine the total power consumption based on the sum of the server energy consumption, the air conditioning energy consumption, and the storage device energy consumption, and determine the electronic equipment energy consumption based on the sum of the server energy consumption and the storage device energy consumption; The second unit is used to determine the power usage efficiency based on the ratio of the energy consumption of the electronic device to the total power consumption; The third unit is configured to determine that a reasonable power consumption state is represented when the power usage efficiency is greater than or equal to an efficiency threshold, and to determine that an unreasonable power consumption state is represented when the power usage efficiency is less than the efficiency threshold.

6. The intelligent energy-saving chassis for a data center according to any one of claims 1 to 4, characterized in that: When the power consumption status representation is unreasonable, determining the predicted server energy consumption, the predicted air conditioning energy consumption, and the predicted storage device energy consumption based on the current energy consumption data and the historical energy consumption data includes: When the power consumption status is not represented reasonably, the current energy consumption data and historical energy consumption data are analyzed through the long short-term memory network to determine the predicted server energy consumption, predicted air conditioning energy consumption and predicted storage device energy consumption. The current energy consumption data and the historical energy consumption data both include the specific time corresponding to the server energy consumption, the air conditioning energy consumption, and the storage device energy consumption.

7. The intelligent energy-saving chassis for a data center according to any one of claims 1 to 4, characterized in that: The adjustment module includes a first adjustment unit, a second adjustment unit, a third adjustment unit, a fourth adjustment unit, a fifth adjustment unit, and a sixth adjustment unit: a first adjustment unit, configured to determine a predicted total power consumption based on the sum of the predicted server energy consumption, the predicted air conditioning energy consumption, and the predicted storage device energy consumption, and to determine a predicted electronic device energy consumption based on the sum of the predicted server energy consumption and the predicted storage device energy consumption; The second adjustment unit is configured to determine a predicted power usage efficiency based on a ratio of the predicted electronic device energy consumption to the predicted total power consumption, and when the predicted power usage efficiency is less than an efficiency threshold, determine the proportion of the predicted server energy consumption, the predicted air conditioning energy consumption, and the predicted storage device energy consumption in the predicted total power consumption, respectively, to obtain a server proportion, an air conditioning energy consumption proportion, and a storage device proportion; a third adjustment unit, configured to analyze the historical energy consumption data to determine a first average proportion of server energy consumption to total power consumption, a second average proportion of air conditioner energy consumption to total power consumption, and a third average proportion of storage device energy consumption to total power consumption; A fourth adjustment unit is configured to reduce the predicted server energy consumption when the server proportion is greater than the first average proportion, so as to make the representation of the power consumption state reasonable; a fifth adjusting unit, configured to reduce the predicted air conditioning energy consumption when the air conditioning energy consumption ratio is greater than the second average ratio, so as to make the representation of the power consumption state reasonable; The sixth adjustment unit is configured to reduce the predicted energy consumption of the storage device when the storage device proportion is greater than the third average proportion, so as to make the representation of the power consumption state reasonable.

8. The intelligent energy-saving chassis for a data center according to claim 7, characterized in that: The specific reductions in predicted server energy consumption, predicted air conditioning energy consumption, and predicted storage device energy consumption are as follows: reducing the frequency and voltage of the chip and setting the state of the idle chip to a sleep state; At least one of increasing the supply air temperature and closing air dampers in non-critical areas; At least one of lowering the RAID level and lowering the OP ratio.

9. The intelligent energy-saving chassis for a data center according to claim 5, characterized in that: The adjustment module includes a seventh adjustment unit, which is configured to: When the predicted power efficiency is lower than the efficiency threshold, the backup power supply is turned on to supply power in conjunction with the main power supply; or, When the main power supply is in an abnormal state, the backup power supply is turned on to supply power.

10. An intelligent energy-saving chassis for a data center, characterized in that: include: A processor and a memory, wherein the memory stores instructions, and the instructions are loaded and executed by the processor to implement the following method: Determine the historical and current energy consumption data of the data center, including server energy consumption, air conditioning energy consumption, and storage device energy consumption; Determine the power consumption status based on current energy consumption data; When the power consumption status is not represented reasonably, the predicted server energy consumption, predicted air conditioning energy consumption and predicted storage device energy consumption are determined based on the current energy consumption data and historical energy consumption data; According to the predicted server energy consumption, the predicted air conditioning energy consumption and the predicted storage device energy consumption, at least one of the server parameters, the air conditioning parameters and the storage device parameters is adjusted to make the power consumption state representation reasonable.

Images