A method and system for data center energy consumption management

Abstract

The application provides a data center energy consumption management method and system, the method comprises the following steps: obtaining historical access traffic data of a data center, training a long short-term memory network based on the historical access traffic data to obtain a traffic prediction model; counting the server types and quantities corresponding to the historical access traffic, and counting the average waiting time delay of users; constructing an influence model of working environment on server performance; predicting the access traffic corresponding to different time periods based on the traffic prediction model, under the condition that the average waiting time delay meets the requirement that it is less than a predetermined threshold, allocating the number of servers corresponding to different time periods, and setting corresponding refrigeration equipment through the influence model with the minimum energy consumption as the target; and based on the actual average waiting time delay of users and the working environment sampling, feedback adjusting the number of servers and the refrigeration equipment. Through the scheme, the energy consumption of the data center can be reduced under the premise of guaranteeing the normal use experience of users.

Description

Technical Field

[0001] This invention relates to the field of computers, and more particularly to a data center energy management method and system. Background Technology

[0002] Data centers typically provide services such as computing, management, transmission, and storage of network data, and contain a large number of server hosts and other equipment. Due to the large number of servers in a data center, the servers consume a significant amount of electricity during operation and the control of server environmental temperature. Under the current requirements of low-carbon development, it is necessary to reduce the energy consumption of data centers.

[0003] Existing energy consumption control methods mostly adjust the energy consumption of the cooling system while ensuring that the host operates at a normal ambient temperature. This can be done by setting the air conditioning temperature and the number of air conditioners turned on. However, these solutions do not take into account the impact of server power consumption, resulting in data center energy consumption remaining high. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide a data center energy consumption management method and system to solve the problem of high energy consumption in existing data centers.

[0005] In a first aspect of the present invention, a data center energy consumption management method is provided, comprising:

[0006] Obtain historical access traffic data from the data center, and train the Long Short Memory Network based on the historical access traffic data to obtain a traffic prediction model;

[0007] Analyze the server types and quantities corresponding to historical access traffic, and calculate the average user wait time.

[0008] Construct a model to illustrate the impact of the working environment on server performance;

[0009] Based on the traffic prediction model, the access traffic corresponding to different time periods is predicted. Under the condition that the average waiting time is less than a predetermined threshold, the number of servers corresponding to different time periods is allocated, and with the goal of minimum energy consumption, the corresponding cooling equipment is set through the influence model.

[0010] Based on actual user average latency and working environment sampling, feedback adjustments are made to the number of servers and cooling equipment.

[0011] In a second aspect of the present invention, a data center energy management system is provided, comprising:

[0012] The model training module is used to acquire historical access traffic data of the data center, and to train the Long Short Memory Network based on the historical access traffic data to obtain the traffic prediction model.

[0013] The data statistics module is used to count the server types and quantities corresponding to historical access traffic, and to count the average user waiting time.

[0014] The model building module is used to build a model of the impact of the working environment on server performance.

[0015] The configuration module is used to predict access traffic corresponding to different time periods based on the traffic prediction model, allocate the number of servers corresponding to different time periods under the condition that the average waiting time is less than a predetermined threshold, and set the corresponding cooling equipment through the influence model with the goal of minimum energy consumption.

[0016] The feedback adjustment module is used to adjust the number of servers and cooling equipment based on the actual average user waiting time and working environment sampling.

[0017] In a third aspect of the present invention, an electronic device is provided, comprising at least a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the method as described in the first aspect of the present invention.

[0018] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method provided in the first aspect of the present invention.

[0019] In this embodiment of the invention, the number of servers is configured according to the changes in data center access traffic, and cooling equipment is set up according to the impact of the working environment on server energy consumption and computing efficiency. This allows for the reasonable setting of the number of servers while ensuring that user services are not affected. Furthermore, the impact of ambient temperature on server power consumption and operating efficiency is considered, and cooling equipment is set up in a reasonable manner. This effectively reduces the power consumption of servers and cooling equipment, thereby effectively reducing the total energy consumption of the data center. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 A flowchart illustrating a data center energy management method according to an embodiment of the present invention;

[0022] Figure 2A schematic diagram of a data center energy management system provided as an embodiment of the present invention;

[0023] Figure 3 This is a schematic diagram of the structure of an electronic device provided in one embodiment of the present invention. Detailed Implementation

[0024] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0025] The terms "comprising" and other similar expressions used in the specification, claims, and accompanying drawings of this invention are intended to cover a non-exclusive inclusion, such as a process, method, system, or apparatus that includes a series of steps or units but is not limited to the listed steps or units. Furthermore, "first" and "second" are used to distinguish different objects and are not intended to describe a specific order.

[0026] See Figure 1 , Figure 1 A flowchart illustrating a data center energy management method provided in this embodiment of the invention includes:

[0027] S101. Obtain historical access traffic data of the data center, and train the Long Short Memory Network based on the historical access traffic data to obtain a traffic prediction model.

[0028] For data center services provided to specific or a subset of customers, server access traffic data for that customer can be obtained. Alternatively, server user access traffic data for a portion or all of the data center can be obtained. For servers identified as providing services to customers, historical access traffic data for these servers at different time periods can be compiled.

[0029] Based on historical traffic data from different time periods, the constructed traffic prediction model can be trained to obtain the trained traffic prediction model, which can be used to predict the access traffic corresponding to different time periods.

[0030] For example, a day's access traffic can be divided into two-hour time periods. By counting the traffic in each time period and training a traffic prediction model, the traffic in the next time period can be predicted.

[0031] S102. Calculate the server type and number corresponding to historical access traffic, and calculate the average user waiting time.

[0032] For historical traffic data across different time periods, it's necessary to obtain the number of servers corresponding to different traffic volumes. For example, if the number of visits exceeds one million, three servers need to be activated. The number of servers activated varies depending on the traffic range. Simultaneously, it's necessary to standardize the user latency corresponding to different numbers of servers activated under varying access traffic conditions.

[0033] This involves calculating the peak and average access volumes for each time period, as well as the corresponding maximum and average number of servers; calculating the peak and average waiting latency for each time period; and constructing a relationship model between data access volume, number of servers, and waiting latency in the data center.

[0034] In some embodiments, when peak latency exceeds a certain value, it is necessary to consider adding additional servers. Alternatively, when peak traffic exceeds a certain value, it is necessary to consider adding additional servers.

[0035] S103. Construct a model to illustrate the impact of the working environment on server performance.

[0036] Servers exhibit varying computing efficiency and power consumption depending on their operating environment. The operating environment generally refers to ambient temperature, but may also include humidity.

[0037] The impact of the working environment on server performance can be determined by obtaining relevant impact data from the server manufacturer, depending on the server model, or by conducting your own tests.

[0038] Among them, the relationship curves between ambient temperature and server computing speed and power consumption were tested; with the optimization objectives of minimizing the power consumption of cooling equipment, minimizing server power consumption, and maximizing server computing speed, an impact model of ambient temperature on server performance was constructed.

[0039] Furthermore, we establish relationship models between cooling equipment power consumption and ambient temperature, ambient temperature and server power consumption, and server computing efficiency, respectively. The first optimization objective is to minimize the total power consumption of the cooling equipment and the server, and the second optimization objective is to maximize the server computing performance. We also set the weights for the optimization objectives.

[0040] S104. Based on the traffic prediction model, predict the access traffic corresponding to different time periods. Under the condition that the average waiting time is less than a predetermined threshold, allocate the number of servers corresponding to different time periods, and set the corresponding cooling equipment through the influence model with the goal of minimum energy consumption.

[0041] Based on traffic prediction data, the corresponding number of servers can be allocated and activated, which can avoid activating too many or too few servers while ensuring user requests are met, and also avoid problems caused by temporarily activating or deactivating servers.

[0042] The cooling equipment generally includes air conditioners, fans, liquid cooling equipment, etc., but may also be other equipment used for server ambient temperature regulation, which is not limited here.

[0043] Given a fixed number of servers, while ensuring the normal operation of the regional servers, it is also necessary to reduce the power consumption of the servers and cooling equipment. That is, to minimize the power consumption of the servers and cooling equipment, the ambient temperature should be set and the corresponding cooling equipment settings should be configured.

[0044] Specifically, to minimize the total power consumption of the cooling equipment and the server, and under the premise that the server's computing performance meets the waiting latency constraint, the corresponding ambient temperature and the number of cooling equipment are determined based on the aforementioned influence model; and the cooling equipment is configured based on the ambient temperature.

[0045] Since server computing efficiency decreases when the ambient temperature is too high, it is necessary to consider the constraints of user waiting latency.

[0046] S105. Based on the actual average user waiting time and working environment sampling, adjust the number of servers and cooling equipment accordingly.

[0047] Real-time monitoring of user latency and sampling of ambient temperature are used to make appropriate adjustments to the server and cooling equipment, preventing excessive user latency due to system errors and the impact of ambient temperature on the normal operation of the server.

[0048] The method provided in this embodiment can effectively reduce the power consumption of servers and cooling equipment while ensuring the normal operation of the data center, thereby reducing the total power consumption of the data center.

[0049] It should be understood that the sequence numbers of the steps in the above embodiments do not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0050] Figure 2 This is a schematic diagram of a data center energy management system provided in an embodiment of the present invention. The system includes:

[0051] The model training module 210 is used to acquire historical access traffic data of the data center, and to train the Long Short Memory Network based on the historical access traffic data to obtain a traffic prediction model.

[0052] The data statistics module 220 is used to count the server types and quantities corresponding to historical access traffic, and to count the average user waiting time.

[0053] The process of statistically analyzing the server types and quantities corresponding to historical access traffic and calculating the average user wait time includes:

[0054] Statistical analysis of peak and average access volumes for each time period, along with the corresponding maximum and average number of servers.

[0055] Calculate the peak waiting time and average waiting time for each time period;

[0056] Construct a relationship model between data access volume, number of servers, and waiting latency in the data center.

[0057] Model building module 230 is used to build an impact model on server performance corresponding to the working environment;

[0058] The impact model on server performance corresponding to the constructed working environment includes:

[0059] The curves showing the relationship between ambient temperature and server processing speed and power consumption were tested.

[0060] With the optimization objectives of minimizing the power consumption of cooling equipment, minimizing the power consumption of servers, and maximizing the server's computing speed, a model is constructed to illustrate the influence of ambient temperature on server performance.

[0061] Furthermore, we establish relationship models between cooling equipment power consumption and ambient temperature, ambient temperature and server power consumption, and server computing efficiency, respectively. The first optimization objective is to minimize the total power consumption of the cooling equipment and the server, and the second optimization objective is to maximize the server computing performance. We also set the weights for the optimization objectives.

[0062] Configuration module 240 is used to predict access traffic corresponding to different time periods based on the traffic prediction model, allocate the number of servers corresponding to different time periods under the condition that the average waiting time is less than a predetermined threshold, and set the corresponding cooling equipment through the influence model with the goal of minimum energy consumption.

[0063] Among them, the total power consumption of the cooling equipment and the server is minimized. Under the premise that the server's computing performance meets the waiting latency constraint, the corresponding ambient temperature and the number of cooling equipment are determined based on the influence model.

[0064] The refrigeration equipment is configured based on the ambient temperature.

[0065] The feedback adjustment module 250 is used to adjust the number of servers and cooling equipment based on the actual average user waiting time and working environment sampling.

[0066] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the systems and modules described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0067] Figure 3This is a schematic diagram of an electronic device according to an embodiment of the present invention. The electronic device is used for energy management in a data center. Figure 3 As shown, the electronic device 3 of this embodiment includes: a memory 310, a processor 320, and a system bus 330. The memory 310 includes an executable program 3101 stored thereon. As those skilled in the art will understand, Figure 3 The electronic device structure shown does not constitute a limitation on the electronic device and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0068] The following is combined with Figure 3 A detailed introduction to each component of the electronic device:

[0069] The memory 310 can be used to store software programs and modules. The processor 320 executes various functional applications and data processing of the electronic device by running the software programs and modules stored in the memory 310. The memory 310 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device (such as cached data), etc. In addition, the memory 310 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0070] The memory 310 contains an executable program 3101 for a sign extraction method. This executable program 3101 can be divided into one or more modules / units, which are stored in the memory 310 and executed by the processor 320 to predict and warn of regional vehicle traffic accident risks. Each module / unit can be a series of computer program instruction segments capable of performing a specific function, describing the execution process of the computer program 3101 in the electronic device 3. For example, the computer program 3101 can be divided into a model training module, a data statistics module, a model building module, a configuration module, and a feedback adjustment module.

[0071] Processor 320 is the control center of the electronic device. It connects various parts of the electronic device via various interfaces and lines. By running or executing software programs and / or modules stored in memory 310, and by calling data stored in memory 310, it performs various functions and processes data, thereby monitoring the overall status of the electronic device. Optionally, processor 320 may include one or more processing units; preferably, processor 320 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, application programs, etc., and the modem processor mainly handles wireless communication. It is understood that the aforementioned modem processor may not be integrated into processor 420.

[0072] The system bus 330 is used to connect various functional components within the computer, transmitting data, address, and control information. Its type can be, for example, a PCI bus, an ISA bus, or a VESA bus. Instructions from the processor 320 are transmitted to the memory 310 via the bus, and the memory 310 sends data back to the processor 320. The system bus 330 is responsible for data and instruction exchange between the processor 320 and the memory 310. Of course, the system bus 330 can also connect to other devices, such as network interfaces and display devices.

[0073] In this embodiment of the invention, the executable program executed by the processing 320 of the electronic device includes:

[0074] Obtain historical access traffic data from the data center, and train the Long Short Memory Network based on the historical access traffic data to obtain a traffic prediction model;

[0075] Analyze the server types and quantities corresponding to historical access traffic, and calculate the average user wait time.

[0076] Construct a model to illustrate the impact of the working environment on server performance;

[0077] Based on the traffic prediction model, the access traffic corresponding to different time periods is predicted. Under the condition that the average waiting time is less than a predetermined threshold, the number of servers corresponding to different time periods is allocated, and with the goal of minimum energy consumption, the corresponding cooling equipment is set through the influence model.

[0078] Based on actual user average latency and working environment sampling, feedback adjustments are made to the number of servers and cooling equipment.

[0079] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0080] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0081] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

[0082] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A data center energy consumption management method, characterized in that, include: Obtain historical access traffic data from the data center, and train the Long Short Memory Network based on the historical access traffic data to obtain a traffic prediction model; The system analyzes the server types and quantities corresponding to historical access traffic and the average user wait time. Specifically, it analyzes the peak access volume, average access volume, maximum number of servers, and average number of servers for each time period. Calculate peak and average latency for each time period; construct a model showing the relationship between data access volume, number of servers, and latency in the data center. Tests were conducted to obtain curves showing the relationship between ambient temperature and server computing speed and power consumption. A model was constructed to demonstrate the impact of ambient temperature on server performance: models were established for the relationship between cooling equipment power consumption and ambient temperature, ambient temperature and server power consumption, and ambient temperature and server computing efficiency. The primary optimization objective was to minimize the total power consumption of the cooling equipment and server, and the secondary optimization objective was to maximize server computing performance. Weights were then assigned to these optimization objectives. Based on the traffic prediction model, the access traffic corresponding to different time periods is predicted. Under the condition that the average waiting latency is less than a predetermined threshold, the number of servers corresponding to different time periods is allocated. With the goal of minimizing energy consumption, the corresponding cooling equipment is set through the influence model. Among them, the total power consumption of cooling equipment and servers is minimized. Under the premise that the server computing performance meets the waiting latency constraint, the corresponding ambient temperature and the number of cooling equipment are determined based on the influence model. The cooling equipment is set based on the ambient temperature. Based on actual user average latency and working environment sampling, feedback adjustments are made to the number of servers and cooling equipment.

2. A data center energy management system, characterized in that, include: The model training module is used to acquire historical access traffic data of the data center, and to train the Long Short Memory Network based on the historical access traffic data to obtain the traffic prediction model. The data statistics module is used to count the server types and numbers corresponding to historical access traffic, and to count the average user waiting time. Specifically, it includes counting the peak access volume, average access volume, and the corresponding maximum number of servers and average number of servers for each time period. Calculate peak and average latency for each time period; construct a model showing the relationship between data access volume, number of servers, and latency in the data center. The model building module is used to test and obtain the relationship curves between ambient temperature and server computing speed and power consumption; to build the impact models of ambient temperature and server performance: to establish the relationship models between cooling equipment power consumption and ambient temperature, ambient temperature and server power consumption, and ambient temperature and server computing efficiency; with the minimum total power consumption of cooling equipment and server as the first optimization objective and the maximum server computing performance as the second optimization objective, and to set the optimization objective weights; The configuration module is used to predict access traffic for different time periods based on the traffic prediction model, allocate a number of servers for different time periods under the condition that the average waiting latency is less than a predetermined threshold, and set the corresponding cooling equipment based on the influence model with the goal of minimizing energy consumption. Specifically, it aims to minimize the total power consumption of cooling equipment and servers. Under the premise that the server computing performance meets the waiting latency constraint, it determines the corresponding ambient temperature and the number of cooling equipment based on the influence model; and sets the cooling equipment based on the ambient temperature. The feedback adjustment module is used to adjust the number of servers and cooling equipment based on the actual average user waiting time and working environment sampling.

3. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the data center energy management method as described in claim 1.

4. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed, it implements the steps of the data center energy management method as described in claim 1.

Images