Server cooling methods, devices, storage media, and server cooling systems

Abstract

This application provides a server cooling method, apparatus, storage medium, and server cooling system. The method includes: acquiring production requirements from the production end; determining the cooling mode of each server in a server cluster configured for the production end based on the production requirements; determining the pre-executed cooling logic for each server based on the cooling mode of each server; and determining whether the currently executed cooling logic is correct based on the difference between the pre-executed cooling logic and the currently executed cooling logic. This solution uses a software comparison method, which is highly efficient and less prone to errors, eliminating the need for manual judgment.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the field of servers, and in particular, to a server refrigeration method and device, a computer readable storage medium, and a server refrigeration system. BACKGROUND

[0002] The heat dissipation problem is an important problem to be solved in server development. On the one hand, it is necessary to reduce energy waste, and on the other hand, it is necessary to ensure normal heat dissipation of the server, so that various devices and external equipment can work normally. At present, the baseboard management controller (BMC) is used to control the fans in the server to solve the heat dissipation problem. The corresponding fan control parameters are different under the air cooling, cold plate and immersion liquid cooling environments.

[0003] The current scheme often uses hardware information (such as GPIO, etc.) to distinguish the refrigeration mode of the server, and this way has a high error rate. SUMMARY

[0004] Embodiments of the present application provide a server refrigeration method, device, computer readable storage medium and server refrigeration system to at least solve the problem that the hardware information (such as GPIO, etc.) is used to distinguish the refrigeration mode of the server in the related art, and this way has a high error rate.

[0005] According to an embodiment of the present application, a server refrigeration method is provided, comprising:

[0006] Obtaining production requirements of a production end; determining refrigeration modes of each server in a server cluster configured for the production end according to the production requirements, the server cluster comprising at least one server, and the refrigeration mode of each server being any one of an air cooling mode, a cold plate mode and an immersion liquid cooling mode;

[0007] Determining refrigeration logic to be executed by each server according to the refrigeration mode of each server, and in the process of refrigeration of each server, obtaining real-time running data of each device in each server in real time, extracting refrigeration logic being executed from the real-time running data, and determining whether the refrigeration logic being executed is correct according to the difference between the refrigeration logic to be executed and the refrigeration logic being executed.

[0008] In an exemplary embodiment, the method further comprises: constructing a mapping relationship between the refrigeration mode and a mode identifier;

[0009] Storing the mode identifier corresponding to the refrigeration mode of each server in a memory of a mainboard of each server;

[0010] The method further comprises:

[0011] According to the mode identifier of each server, the corresponding refrigeration logic is called to realize refrigeration.

[0012] In one example embodiment, the method further comprises:

[0013] In the case where the executing refrigeration logic is incorrect, prompt information is generated;

[0014] The prompt information is sent to an operation and maintenance end, and the prompt information is used to prompt an operation and maintenance personnel to perform secondary configuration of the refrigeration logic of the server.

[0015] In one example embodiment, the method further comprises:

[0016] In the process of refrigeration of each server, if a target server works abnormally, the refrigeration mode of the target server is acquired;

[0017] According to the refrigeration mode of the target server, a substitute server is selected, and the substitute server is used to replace the target server, wherein the refrigeration mode of the target server is the same as the refrigeration mode of the substitute server.

[0018] In one example embodiment, the method further comprises:

[0019] In the process of refrigeration of each server, if a target server works abnormally, an abnormal system event log is generated.

[0020] In one example embodiment, the pre-executed refrigeration logic of each server is determined according to the refrigeration mode of each server, and the pre-executed refrigeration logic comprises one of the following:

[0021] In the case where the refrigeration mode of the server is the air-cooled mode, the pre-executed refrigeration logic is linear control according to the temperature value of the temperature control point in the server, and the PWM value of the control fan speed is determined;

[0022] In the case where the refrigeration mode of the server is the cold plate mode, the pre-executed refrigeration logic is to control the refrigerant to flow from a first area to a second area;

[0023] In the case where the refrigeration mode of the server is the immersion liquid cooling mode, the pre-executed refrigeration logic is to disable all fan sensors.

[0024] According to another embodiment of the application, a server refrigeration device is provided, comprising:

[0025] A first acquisition unit is configured to acquire production demand of a production end.

[0026] The first determining unit is configured to determine a cooling mode of each server in a server cluster configured for the production end according to the production demand, the server cluster comprising at least one server, and the cooling mode of each server being any one of a wind cooling mode, a cold plate mode, and an immersion liquid cooling mode.

[0027] The second determining unit is configured to determine a pre-executed cooling logic of each server according to the cooling mode of each server, and to acquire real-time running data of each device in each server in real time during the cooling of each server, extract an executing cooling logic from the real-time running data, and determine whether the executing cooling logic is correct according to a difference between the pre-executed cooling logic and the executing cooling logic.

[0028] According to still another embodiment of the present application, a computer readable storage medium is provided, and the computer readable storage medium stores a computer program. When the computer program is executed, the steps in any one of the method embodiments described above are performed.

[0029] According to still another embodiment of the present application, a server cooling system is provided, and the server cooling system comprises a baseboard management controller configured to perform the steps of any one of the methods described above, and a server cluster comprising at least one server and in communication with the baseboard management controller.

[0030] In an example embodiment, the server cooling system further comprises an external memory, and the baseboard management controller has a first interface through which the baseboard management controller is connected to the external memory, and the external memory is configured to store a cooling logic.

[0031] In an example embodiment, the baseboard management controller further has at least one second interface, and the baseboard management controller is connected to one of the servers through one of the second interfaces.

[0032] According to the application, firstly, the production demand of the production end is acquired; then, the cooling mode of each server in the server cluster configured for the production end is determined according to the production demand, the server cluster comprises at least one server, and the cooling mode is any one of the air cooling mode, the cold plate mode and the immersion liquid cooling mode; finally, the pre-executed cooling logic of each server is determined according to the cooling mode of each server. During the cooling process of each server, the real-time running data of each device in each server is acquired in real time, the pre-executed cooling logic is extracted from the real-time running data, and whether the pre-executed cooling logic is correct is determined according to the difference between the pre-executed cooling logic and the pre-executed cooling logic. The cooling mode of each server in the server cluster is determined according to the demand of the production end, and the pre-executed cooling logic is extracted from the real-time running data of each device in each server, and then whether the pre-executed cooling logic is correct is determined by comparing the difference between the pre-executed cooling logic and the pre-executed cooling logic. Compared with the method of distinguishing the cooling mode of the server by using hardware information (such as GPIO), the application adopts a software comparison method, which is more efficient and less prone to errors, and does not need to be judged by human. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 is a flowchart of a server cooling method according to an embodiment of the application;

[0034] Figure 2 is a schematic diagram of a server cooling device according to an embodiment of the application;

[0035] Figure 3 is a schematic diagram of a server cooling system according to an embodiment of the application.

[0036] Among them, the above-mentioned drawings include the following reference signs:

[0037] 10, baseboard management controller; 11, first interface; 12, second interface; 20, server; 30, external memory. DETAILED DESCRIPTION

[0038] Hereinafter, the embodiments of the application will be described in detail with reference to the accompanying drawings and in conjunction with the embodiments.

[0039] It should be noted that the terms "first", "second" and the like in the specification and claims of the application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence.

[0040] In order to facilitate description, the following describes some nouns or terms related to the embodiments of the application:

[0041] Air cooling mode: is installed in the server by the fan device, by the fan device rotation for server temperature adjustment;

[0042] Cold plate mode: also known as indirect liquid cooling technology. This technology directly guides the coolant to the heat source, and because the specific heat of liquid is larger than that of air, the heat dissipation speed is much higher than that of air, so the refrigeration efficiency is much higher than that of air cooling heat dissipation;

[0043] Immersion liquid cooling mode: The immersion liquid cooling environment is to completely immerse the server mainboard, CPU, memory and other high-heat components in the refrigerant. In the working state, each heat generating component will generate heat, causing the refrigerant temperature to rise. When the refrigerant temperature rises to the boiling point corresponding to the system pressure, the refrigerant working medium changes from liquid to gas, and the heat is transferred through the vaporization heat absorption.

[0044] Baseboard Management Controller (BMC): can perform firmware upgrade, view machine equipment and other operations on the machine in the state of the machine not being started.

[0045] Intelligent Platform Management Interface (IPMI): IPMI can cross different operating systems, firmware and hardware platforms

[0046] Redfish interface: is a management standard based on HTTPs service, which realizes device management by using RESTful interface. Each HTTPs operation is in the form of UTF-8 encoded JSON, which submits or returns a resource. Just like a web application returns HTML to a browser, a RESTful interface returns data to a client in the form of JSON through the same transmission mechanism (HTTPs).

[0047] A method running on a baseboard management controller (BMC) is provided in the embodiment, Figure 1 According to the server refrigeration method flowchart of the embodiment of the present application, as Figure 1 The flowchart includes the following steps:

[0048] Step S102, obtaining the production demand of the production end;

[0049] Specifically, the production demand of the production end can include: the type of the product produced, the quantity of the product produced, the speed of product production, etc. The specific production demand is determined according to the specific situation of the production end;

[0050] Step S104, determine the cooling mode of each server in the server cluster configured for the production end according to the production requirements. The server cluster includes at least one server, and the cooling mode of each server is any one of air cooling mode, cold plate mode and immersion liquid cooling mode.

[0051] Of course, as technology develops, the cooling mode can also be a mode other than air cooling mode, cold plate mode and immersion liquid cooling mode. The corresponding cooling logic can be set according to the new mode;

[0052] For example, according to the production requirements, it is determined that 100 servers are needed, of which 20 servers are set to air cooling mode, 30 servers are set to cold plate mode, and 50 servers are set to immersion liquid cooling mode.

[0053] For example, according to the production requirements, it is determined that 100 servers are needed, of which 50 servers are set to air cooling mode and 50 servers are set to cold plate mode. The actual requirements can be set according to the actual requirements.

[0054] Step S106, determine the pre-executed cooling logic of each server according to the cooling mode of each server, and in the process of cooling each server, real-time running data of each device in each server is obtained in real time, the executing cooling logic is extracted from the real-time running data, and whether the executing cooling logic is correct is determined according to the difference between the pre-executed cooling logic and the executing cooling logic.

[0055] The difference between the pre-executed cooling logic and the executing cooling logic can be obtained by comparing the difference in numerical value, the difference in data fluctuation, etc., wherein the difference in numerical value can be the difference in maximum value, the difference in minimum value, the difference in average value, etc.

[0056] In the process of cooling each server, real-time running data of each device in each server is obtained in real time; the real-time running data reflects the cooling process being performed.

[0057] Compare the pre-executed cooling logic and the real-time running data to obtain a comparison result; determine whether the executing cooling logic is correct according to the comparison result. By comparing the cooling logic corresponding to the mode identifier stored in the memory of the mainboard of the server and the real-time running data, it can be determined whether the cooling logic being performed is correct; in this way, if the thermal environment control is incorrect during production, it can be identified in time to ensure correct heat dissipation.

[0058] Of course, in some cases, if the difference between the pre-executed refrigeration logic and the refrigeration logic being executed is small enough to affect the refrigeration effect, the refrigeration logic can also not be reconfigured.

[0059] Of course, the adjustment strategy corresponding to the difference between the different pre-executed refrigeration logic and the refrigeration logic being executed can also be stored in the memory for subsequent use in the refrigeration process to improve efficiency and save resources.

[0060] Specifically, the correspondence between the difference between the pre-executed refrigeration logic and the refrigeration logic being executed and the corresponding adjustment strategy can be stored in a database table for subsequent calling. Of course, different correspondences are stored in different positions in the database table, and the position information is retained for subsequent use.

[0061] Specifically, one refrigeration mode corresponds to one refrigeration logic, for example:

[0062] In the case where the refrigeration mode of the server is the air cooling mode, the pre-executed refrigeration logic is to perform linear control according to the temperature value of the temperature control point in the server, and determine the PWM value of the fan speed;

[0063] The PID algorithm can be used to calculate the PWM value of the fan speed;

[0064] In the case where the refrigeration mode of the server is the cold plate mode, the pre-executed refrigeration logic is to control the refrigerant to flow from the first region to the second region;

[0065] In the cold plate environment, the working fluid is used as an intermediate heat transfer medium to transfer heat from the hot area to a remote place for cooling. For such a server, an air-cooled hybrid mode is generally used, and the BMC needs to cancel the control of part of the fan and disable the corresponding fan sensor.

[0066] In the case where the refrigeration mode of the server is the immersion liquid cooling mode, the pre-executed refrigeration logic is to disable all fan sensors. That is, the immersion liquid cooling mode has no fan, and the server mainboard does not need to be configured with a fan when the whole machine is replaced. That is, the BMC needs to cancel all fan controls and disable all fan sensors in the immersion liquid cooling environment.

[0067] The server refrigeration method of the present application first acquires the production demand of the production end, then determines the refrigeration mode of each server in the server cluster configured for the production end according to the production demand, the server cluster includes at least one server, and the refrigeration mode of each server is any one of air cooling mode, cold plate mode and immersion liquid cooling mode, and finally determines the pre-executed refrigeration logic of each server according to the refrigeration mode of each server. During the refrigeration process of each server, the real-time running data of each device in each server is acquired in real time, the executing refrigeration logic is extracted from the real-time running data, and whether the executing refrigeration logic is correct is determined according to the difference between the pre-executed refrigeration logic and the executing refrigeration logic. The refrigeration mode of each server in the server cluster is determined according to the demand of the production end, the executing refrigeration logic is extracted from the real-time running data of each device in each server, and then whether the executing refrigeration logic is correct is determined by comparing the difference between the pre-executed refrigeration logic and the executing refrigeration logic. Compared with the method of distinguishing the refrigeration mode of the server by using hardware information (such as GPIO, etc.), the present application adopts a software comparison method, which is more efficient and less prone to errors, and does not require human judgment.

[0068] In order to accurately obtain the refrigeration mode of each server during refrigeration, the method further comprises:

[0069] mapping relationship between the refrigeration mode and the mode identifier is constructed;

[0070] For example, the setting options are: 0x00 represents air cooling mode, 0x01 represents cold plate mode, and 0x02 represents cold plate immersion liquid cooling mode, and this mapping relationship is configured during installation, and is 0xFF during production.

[0071] For example, the setting options are: 0x01 represents air cooling mode, 0x02 represents cold plate mode, and 0x03 represents cold plate immersion liquid cooling mode, and this mapping relationship is configured during installation, and is 0xFF during production, which can be set according to actual demand.

[0072] The method further comprises: according to the mode identifier corresponding to each server, the corresponding refrigeration logic is called to realize refrigeration. Facilitate refrigeration.

[0073] The mode identifier corresponding to the refrigeration mode of each server is stored in the memory of the mainboard of each server. The mode identifier is stored in the memory of the mainboard of each server, which facilitates the subsequent baseboard management controller (BMC) to call when performing refrigeration control.

[0074] Specifically, the BMC provides an external interface, so that the production end stores the Thermal information (i.e., mode identifier) in the non-volatile storage medium of the motherboard during production, and the BMC performs different heat dissipation control logic according to the information during the running stage. In this way, the hardware information (such as GPIO, etc.) is no longer used for differentiation, which saves hardware resources and is more flexible. The same motherboard can be installed in different Thermal environments, increasing the utilization rate of the motherboard.

[0075] The external interface provided by the BMC can be an IPMI Command, a Redfish interface, etc. The BMC obtains the current Thermal configuration from the non-volatile storage medium and performs different heat dissipation control strategies.

[0076] To achieve accurate refrigeration, further, after determining whether the refrigeration logic being executed is correct according to the comparison result, the method further includes:

[0077] In the case where the refrigeration logic being executed is incorrect, a prompt information is generated;

[0078] Specifically, the prompt information can be a prompted text, a prompted voice, etc.

[0079] Of course, the prompt information can also be displayed on the display screen to timely remind the operation and maintenance personnel. More specifically, a partition display mode can be used, for example, the prompt information corresponding to 1-20 servers is displayed in the first area, the prompt information corresponding to 1-40 servers is displayed in the first area, and the prompt information corresponding to 1-60 servers is displayed in the first area, so as to remind the corresponding operation and maintenance personnel to quickly locate the corresponding abnormal server.

[0080] In addition, the incorrect refrigeration logic (i.e., refrigeration logic anomaly) can be divided into multiple levels, such as the number level of the fan and the speed level of the fan. Therefore, the refrigeration logic anomaly can be divided according to the level and then displayed in different display areas, for example, the highest level is displayed in the uppermost area of the display screen, and the lowest level is displayed in the uppermost area of the display screen, so as to prompt the operation and maintenance personnel to reconfigure according to the level of the refrigeration logic anomaly, thereby saving time and ensuring the refrigeration effect.

[0081] Specifically, the prompt information includes the ID of the server that needs to be configured again, so that the operation and maintenance personnel can find the corresponding server according to the ID of the server and perform secondary configuration on the corresponding server. Compared with the case of not correcting errors in the prior art, the scheme of the present application greatly facilitates the operation of the operation and maintenance personnel, and the operation and maintenance personnel can find and handle problems in time after reconfiguration, thereby stopping loss in time.

[0082] The prompt information is sent to the operation and maintenance end, and the prompt information is used to prompt the operation and maintenance personnel to perform secondary configuration on the refrigeration logic of the server. For example, the logic that should run in air cooling mode actually runs in cold plate mode, the logic that should run in cold plate mode actually runs in air cooling mode, the logic that should run in immersion liquid cooling mode actually runs in air cooling mode, and the like. In the case of reconfiguration.

[0083] For example, when the current Thermal configuration is obtained as an immersion liquid cooling environment, when the fan Present state is detected as Yes, and the Board ID detection is not the immersion corresponding ID, an alarm needs to be given to prompt that the Thermal environment does not match.

[0084] For example, if the BMC obtains the current Thermal configuration as a cold plate environment, when the liquid leakage detection line is not in place or is disconnected, an alarm needs to be given to prompt that the Thermal environment does not match.

[0085] For example, in the case where the current Thermal configuration is obtained as an air cooling mode, the running speed of the fan is detected as a first running speed, but the pre-set running speed is the first running speed. At this time, it is determined that the refrigeration logic is incorrect. One case is that the pre-set refrigeration logic is incorrect, and the other case is that the executing refrigeration logic is incorrect. Whether the pre-set refrigeration logic is incorrect or the executing refrigeration logic needs to be determined according to the real-time refrigeration effect. If the real-time refrigeration effect has reached the expected effect, it proves that the speed of the fan in the executing refrigeration logic is a correct setting, and the speed of the fan in the pre-set refrigeration logic needs to be modified.

[0086] Of course, the running speed of the fan is only exemplary, and the running parameters of the hardware related to the air cooling mode, the cold plate mode, and the immersion liquid cooling mode are within the protection scope of the present application.

[0087] At the same time, for the production stage, the same mainboard does not know what the subsequent configuration (air cooling mode, water cooling mode, immersion liquid cooling mode) is. If the configuration is set in advance, subsequent online reconfiguration or problems in the production stage will cause the Thermal real environment to be different from the set one, and the heat dissipation regulation will be chaotic, which will cause serious problems. In order to ensure the normal production of the production end, the above method further includes:

[0088] In the process of refrigeration of each of the above servers, if the target server works abnormally, the refrigeration mode of the target server is obtained.

[0089] According to the above target server cooling mode selection alternative server, the above-mentioned alternative server is used to replace the above-mentioned target server, wherein the cooling mode of the target server is the same as the cooling mode of the alternative server. For example, if a server in air cooling mode fails, a new server in air cooling mode is used to replace it, if a server in cold plate mode fails, a new server in cold plate mode is used to replace it, if a server in immersion liquid cooling mode fails, a new server in immersion liquid cooling mode is used to replace it, so that the servers with different cooling modes are not switched to affect normal production. The configuration Thermal information of the online reconfigured mainboard does not match the real hardware information, which can be timely distinguished to ensure correct heat dissipation.

[0090] Further, the above method further comprises:

[0091] In the process of cooling each of the above-mentioned servers, if the target server works abnormally, an abnormal system event log is generated.

[0092] In an example embodiment, a display page is also configured, which can be a WEB page, and the abnormal system event log is displayed in the display page when the target server works abnormally.

[0093] Further, the display page can also display the mode identifier corresponding to the cooling mode of each of the servers.

[0094] In this paper, the OEM (manufacturer-defined) SEL (system event log) is used for alarm recording, which does not depend on the addition of sensors, saving resources, because there are many components that need to be monitored in the server scene, so many sensors are needed for adaptation and monitoring, but the number of sensors is limited and cannot be increased indefinitely.

[0095] Through the scheme proposed in this paper, different Thermal environments can be adapted through BMC, without the need for additional hardware resources for identification. BMC can adjust the corresponding heat dissipation strategy according to different Thermal configurations; at the same time, if the configuration is wrong during production or the server is reconfigured in the customer's machine room, causing the settings to mismatch the real hardware, BMC can automatically monitor and alarm, prompting the operation and maintenance personnel to modify in time, and configure the correct Thermal environment.

[0096] In addition, the scheme of the present application can be extended to the operation of the setting type, further verifying whether the setting is correct according to the known environment, and if not, alarming and prompting the user to modify the setting again.

[0097] The execution subject of the above steps can be a substrate management controller, but is not limited thereto.

[0098] From the above description of the embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be implemented by means of software on a general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as a ROM / RAM, a magnetic disk, or an optical disk) and includes a plurality of instructions for causing a terminal device (which can be a mobile phone, a computer, a server, or a network device) to execute the methods described in the various embodiments of the present application.

[0099] In the present embodiment, a server cooling device is also provided, which is used to implement the above embodiments and preferred embodiments, and will not be described again. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, implementation of hardware, or a combination of software and hardware, is also possible and contemplated.

[0100] Figure 2 is a structural block diagram of a server cooling device according to an embodiment of the present application, as shown in Figure 2 , the device includes:

[0101] The first acquisition unit 21 is configured to acquire the production demand of the production end.

[0102] Specifically, the production demand of the production end can include the type of the product to be produced, the quantity of the product to be produced, the speed of the product to be produced, and the like, and the specific production demand is determined according to the specific situation of the production end.

[0103] The first determination unit 22 is configured to determine the cooling mode of each server in the server cluster configured for the production end according to the production demand, wherein the server cluster includes at least one server, and the cooling mode of each server is any one of a wind cooling mode, a cold plate mode, and an immersion liquid cooling mode.

[0104] Of course, as technology develops, the cooling mode can also be a mode other than the wind cooling mode, the cold plate mode, and the immersion liquid cooling mode, and the corresponding cooling logic can be set according to the new mode;

[0105] For example, according to the production demand, it is determined that 100 servers are needed, of which 20 servers are set to the wind cooling mode, 30 servers are set to the cold plate mode, and 50 servers are set to the immersion liquid cooling mode.

[0106] Specifically, for example: according to the production demand, it is determined that 100 servers are needed, of which 50 servers are set to air cooling mode and 50 servers are set to cold plate mode, which can be set according to actual demand;

[0107] The second determination unit 23 is configured to determine the pre-executed refrigeration logic of each server according to the refrigeration mode of each server, and in the process of refrigeration of each server, real-time running data of each device in each server is acquired in real time, the pre-executed refrigeration logic is extracted from the real-time running data, and whether the executed refrigeration logic is correct is determined according to the difference between the pre-executed refrigeration logic and the executed refrigeration logic.

[0108] The difference between the pre-executed refrigeration logic and the executed refrigeration logic can be obtained by comparing the difference in numerical value, the difference in data fluctuation, etc., wherein the difference in numerical value can be the difference in maximum value, the difference in minimum value, the difference in average value, etc. of a parameter in the refrigeration logic.

[0109] Specifically, one refrigeration mode corresponds to one refrigeration logic, for example:

[0110] In the case where the refrigeration mode of the server is the air cooling mode, the pre-executed refrigeration logic is: linear control according to the temperature value of the temperature control point in the server, and determining the PWM value of the fan speed;

[0111] The PID algorithm can be used to calculate the PWM value of the fan speed;

[0112] In the case where the refrigeration mode of the server is the cold plate mode, the pre-executed refrigeration logic is: controlling the refrigerant to flow from the first area to the second area;

[0113] In the cold plate environment, the working fluid is used as an intermediate heat transfer medium to transfer heat from the hot area to a remote place for cooling. For such a server, the air cooling mixed mode is generally used, and the BMC needs to cancel the control of part of the fans and disable the corresponding fan sensors;

[0114] In the case where the refrigeration mode of the server is the immersion liquid cooling mode, the pre-executed refrigeration logic is: disabling all fan sensors. That is, the immersion liquid cooling mode has no fan, and there is no need to configure a fan for the server mainboard when the whole machine is replaced. That is, the BMC needs to cancel all fan controls and disable all fan sensors in the immersion liquid cooling environment.

[0115] The server refrigeration device of the application, the first acquisition unit acquires the production demand of the production end, the first determination unit determines the refrigeration mode of each server in the server cluster configured for the production end according to the production demand, the server cluster includes at least one server, and the refrigeration mode is any one of air cooling mode, cold plate mode and immersion liquid cooling mode; the second determination unit determines the refrigeration logic to be executed by each server according to the refrigeration mode of each server, and in the process of refrigeration of each server, the real-time running data of each device in each server is acquired in real time, the refrigeration logic being executed is extracted from the real-time running data, and whether the refrigeration logic being executed is correct is determined according to the difference between the pre-executed refrigeration logic and the refrigeration logic being executed. The refrigeration mode of each server in the server cluster is determined according to the demand of the production end, and the refrigeration logic being executed is extracted from the real-time running data of each device in each server, and then whether the refrigeration logic being executed is correct is determined by comparing the difference between the pre-executed refrigeration logic and the refrigeration logic being executed. Compared with the method of distinguishing the refrigeration mode of the server by using hardware information (such as GPIO, etc.), the application adopts a software comparison method, which is more efficient and less prone to errors, and does not need to be judged by human.

[0116] Further, the device further comprises a construction unit and a storage unit, the construction unit is used for constructing the mapping relationship between the refrigeration mode and the mode identifier; the storage unit is used for storing the mode identifier corresponding to the refrigeration mode of each server in the memory of the mainboard of each server. For example, the setting options are: 0x00 represents air cooling mode, 0x01 represents cold plate mode, and 0x02 represents cold plate immersion liquid cooling mode, and this mapping relationship is configured during installation, and is 0xFF during production;

[0117] For example, the setting options are: 0x01 represents air cooling mode, 0x02 represents cold plate mode, and 0x03 represents cold plate immersion liquid cooling mode, and this mapping relationship is configured during installation, and is 0xFF during production, which can be set according to actual demand.

[0118] The mode identifier corresponding to the refrigeration mode of each server is stored in the memory of the mainboard of each server. The mode identifier is stored in the memory of the mainboard of each server, which is convenient for subsequent baseboard management controller (BMC) to call when refrigeration control is performed.

[0119] Specifically, the BMC provides an external interface, so that the production end stores the Thermal information (i.e., mode identifier) in the non-volatile storage medium of the motherboard during production, and the BMC performs different heat dissipation control logic in the running stage through the information. In this way, hardware information (such as GPIO, etc.) is no longer used for differentiation, which saves hardware resources and is more flexible. The same motherboard can be installed in different Thermal environments, increasing the utilization rate of the motherboard.

[0120] The external interface provided by the BMC can be: IPMI Command, WEB page option, Redfish interface, etc. The BMC obtains the current Thermal configuration from the non-volatile storage medium and performs different heat dissipation control strategies.

[0121] Further, the above device further includes a retrieval unit for retrieving the corresponding refrigeration logic according to the mode identifier corresponding to each server to realize refrigeration.

[0122] Further, the above device further includes a first generation unit and a sending unit. The first generation unit is used to generate a prompt information in the case that the above-mentioned executing refrigeration logic is incorrect. Specifically, the prompt information can be a prompted text, a prompted voice, etc.

[0123] The sending unit is used to send the above-mentioned prompt information to the operation and maintenance end. The above-mentioned prompt information is used to prompt the operation and maintenance personnel to perform secondary configuration on the refrigeration logic of the server. For example, the logic that should be running is the air cooling mode, but the actual running is the cold plate mode. The logic that should be running is the cold plate mode, but the actual running is the air cooling mode. The logic that should be running is the immersion liquid cooling mode, but the actual running is the air cooling mode, and so on. In this case, re-configuration is required.

[0124] For example, when the current Thermal configuration is obtained as an immersion liquid cooling environment, and it is detected that the fan Present state is Yes and the Board ID detection is not the ID corresponding to the immersion, then an alarm is required to prompt that the Thermal environment is not matched.

[0125] For example, if the BMC obtains the current Thermal configuration as a cold plate environment, and it is detected that the liquid leakage detection line is not in place or is disconnected, then an alarm is required to prompt that the Thermal environment is not matched.

[0126] Meanwhile, for the production stage, the same mainboard does not know what the subsequent configuration (air cooling mode, water cooling mode, and immersion liquid cooling mode) is, and if the configuration is set in advance, subsequent online reconfiguration or problems in the production stage will cause the Thermal real environment to be different from the set environment, and the heat dissipation regulation will be chaotic, resulting in serious problems. To ensure normal production at the production end, further, the above device further includes a third acquisition unit and a selection unit. The third acquisition unit is configured to acquire the cooling mode of a target server if the target server works abnormally during the cooling process of each server. The selection unit is configured to select a replacement server according to the cooling mode of the target server. The replacement server is configured to replace the target server, and the cooling mode of the target server is the same as that of the replacement server. For example, if a server in air cooling mode fails, a new server in air cooling mode is used to replace it. If a server in cold plate mode fails, a new server in cold plate mode is used to replace it. If a server in immersion liquid cooling mode fails, a new server in immersion liquid cooling mode is used to replace it. In this way, different cooling mode servers will not be switched to affect normal production. When the configuration Thermal information of the mainboard after online reconfiguration does not match the real hardware information, it can be timely identified to ensure correct heat dissipation.

[0127] Further, the above device further includes a second generation unit. The second generation unit is configured to generate an abnormal system event log if a target server works abnormally during the cooling process of each server.

[0128] In this paper, the OEM (manufacturer-defined) SEL (system event log) is used for alarm recording. This event does not depend on the addition of sensors, saving resources. Because there are many components that need to be monitored in the server scene, many sensors are needed for adaptation and monitoring, but the number of sensors is limited and cannot be increased indefinitely.

[0129] It should be noted that each of the above modules can be implemented by software or hardware. For the latter, the following implementation manners can be used, but are not limited thereto: all the above modules are located in the same processor; or each of the above modules is located in a different processor in any combination.

[0130] Embodiments of the present application also provide a computer readable storage medium, which stores a computer program. The computer program is configured to execute the steps in any of the above method embodiments when running.

[0131] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard disk, magnetic disk, or optical disk.

[0132] Embodiments of this application also provide a server cooling system, such as Figure 3 As shown, it includes:

[0133] The baseboard management controller 10 is used to perform the steps of the server cooling method described above;

[0134] The server cluster includes at least one server 20 that communicates with the baseboard management controller.

[0135] The baseboard management controller in the server cooling system of this application acquires the production requirements from the production end; determines the cooling mode for each server in the server cluster configured for the production end based on the production requirements, wherein the server cluster includes at least one server, and the cooling mode is any one of air cooling, cold plate cooling, and immersion liquid cooling; and determines the pre-executed cooling logic for each of the servers based on the cooling mode of each server. This achieves the determination of the cooling mode of each server in the server cluster based on the production end's requirements, and the determination of the cooling logic based on the cooling mode of each server, thereby realizing server cooling. Compared to methods that use hardware information (such as GPIO) to distinguish the server's cooling mode, this method is more efficient and less prone to errors.

[0136] Furthermore, such as Figure 3 As shown, the server cooling system also includes an external memory 30. The baseboard management controller 10 has a first interface 11, and the baseboard management controller 10 is connected to the external memory 30 through the first interface 11. The external memory 30 is used to store cooling logic. The cooling logic is stored in the external memory, and the baseboard management controller retrieves the cooling logic from the external memory through the first interface for cooling control.

[0137] Furthermore, such as Figure 3 As shown, the baseboard management controller also has at least one second interface 12, through which the baseboard management controller 10 is connected to one of the servers 20.

[0138] Additionally, the baseboard management controller can also be connected to a server via a second interface, a switch, and a connection port. The switch can select which server's connection port the second interface is successfully connected to.

[0139] More specifically, the second interface is one of: a standard bus interface, a standard PCIe interface, a Redfish interface.

[0140] Embodiments of the present application also provide an electronic device, comprising a memory and a processor, the memory storing a computer program, and the processor being configured to execute the computer program to perform the steps in any of the above method embodiments.

[0141] In an example embodiment, the electronic device can further comprise a transmission device connected to the processor, and an input / output device connected to the processor.

[0142] Specific examples in the embodiments can refer to the examples described in the above embodiments and example implementations, which will not be repeated here.

[0143] Obviously, those skilled in the art should understand that the modules or steps of the present application described above can be realized by general computing devices, which can be concentrated on a single computing device or distributed on a network composed of multiple computing devices, and can be realized by program codes executable by the computing devices, so that they can be stored in storage devices and executed by the computing devices, and in some cases, the steps shown or described can be executed in different order, or they can be manufactured into individual integrated circuit modules, or multiple modules or steps can be manufactured into a single integrated circuit module. Thus, the present application is not limited to any specific combination of hardware and software.

[0144] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:

[0145] 1) The server refrigeration method of the present application, first, the production demand of the production end is obtained; then, the refrigeration mode of each server in the server cluster configured for the production end according to the above production demand is determined, the server cluster includes at least one server, and the refrigeration mode of each server is any one of air cooling mode, cold plate mode and immersion liquid cooling mode; finally, the refrigeration logic to be executed by each server is determined according to the refrigeration mode of each server. And in the process of refrigeration of each server, the real-time running data of each device in each server is obtained in real time, the refrigeration logic being executed is extracted from the real-time running data, and whether the refrigeration logic being executed is correct is determined according to the difference between the refrigeration logic to be executed and the refrigeration logic being executed. The refrigeration mode of each server in the server cluster is determined according to the demand of the production end, and the refrigeration logic being executed is extracted from the real-time running data of each device in each server, and then whether the refrigeration logic being executed is correct is determined by comparing the difference between the refrigeration logic to be executed and the refrigeration logic being executed. Compared with the method of distinguishing the refrigeration mode of the server by using hardware information (such as GPIO, etc.), the present application adopts a software comparison method, which is more efficient and less prone to errors, and does not need human judgment.

[0146] 2) The server refrigeration device of the present application, the first acquisition unit acquires the production demand of the production end, the first determination unit determines the refrigeration mode of each server in the server cluster configured for the production end according to the production demand, the server cluster includes at least one server, and the refrigeration mode is any one of air cooling mode, cold plate mode and immersion liquid cooling mode; the second determination unit determines the pre-executed refrigeration logic of each server according to the refrigeration mode of each server, and in the process of refrigeration of each server, the real-time running data of each device in each server is acquired in real time, the executing refrigeration logic is extracted from the real-time running data, and whether the executing refrigeration logic is correct is determined according to the difference between the pre-executed refrigeration logic and the executing refrigeration logic. The refrigeration mode of each server in the server cluster is determined according to the demand of the production end, and the executing refrigeration logic is extracted from the real-time running data of each device in each server, and then whether the executing refrigeration logic is correct is determined by comparing the difference between the pre-executed refrigeration logic and the executing refrigeration logic. Compared with the method of distinguishing the refrigeration mode of the server by using hardware information (such as GPIO, etc.), the present application adopts a software comparison method, which is more efficient and less prone to errors, and does not need human judgment. The above is only the preferred embodiment of the present application and is not used to limit the present application. For those skilled in the art, the present application can have various changes and variations. Any modification, equivalent replacement, improvement, etc. within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method of refrigeration for a server, the method comprising: The method comprises: obtaining production demand of a production end; determining a cooling mode of each server in a server cluster configured for the production end according to the production demand, the server cluster comprising at least one server, and the cooling mode of each server being any one of a wind cooling mode, a cold plate mode, and an immersion liquid cooling mode; determining a pre-executed cooling logic of each server according to the cooling mode of each server, and in a process of cooling of each server, obtaining real-time running data of each device in each server, extracting an executing cooling logic from the real-time running data, and determining whether the executing cooling logic is correct according to a difference between the pre-executed cooling logic and the executing cooling logic.

2. The method of claim 1, wherein, The method further comprises: building a mapping relationship between the cooling mode and a mode identifier; storing the mode identifier corresponding to the cooling mode of each server in a memory of a mainboard of each server; The method further comprises: according to the mode identifier corresponding to each server, calling the corresponding cooling logic to realize cooling.

3. The method of claim 1, wherein, The method further comprises: in a case where the executing cooling logic is incorrect, generating prompt information; sending the prompt information to an operation and maintenance end, the prompt information being used to prompt an operation and maintenance personnel to perform secondary configuration on the cooling logic of the server.

4. The method of claim 1, wherein, The method further comprises: in the process of cooling of each server, if a target server is abnormal, obtaining a cooling mode of the target server; according to the cooling mode of the target server, selecting a replacement server, the replacement server being used to replace the target server, wherein the cooling mode of the target server is the same as that of the replacement server.

5. The method of claim 1, wherein, The method further comprises: in the process of cooling of each server, if a target server is abnormal, generating an abnormal system event log.

6. The method according to any one of claims 1 to 5, wherein determining the pre-executed cooling logic of each server according to the cooling mode of each server comprises one of the following: in a case where the cooling mode of the server is the wind cooling mode, the pre-executed cooling logic is linear control according to a temperature value of a temperature control point in the server, and determining a PWM value of a fan speed; in a case where the cooling mode of the server is the cold plate mode, the pre-executed cooling logic is controlling refrigerant to flow from a first area to a second area; in a case where the cooling mode of the server is the immersion liquid cooling mode, the pre-executed cooling logic is disabling all fan sensors.

7. A server refrigeration apparatus, characterized by, The method comprises: a first obtaining unit configured to obtain production demand of a production end; a first determining unit configured to determine a cooling mode of each server in a server cluster configured for the production end according to the production demand, the server cluster comprising at least one server, and the cooling mode of each server being any one of a wind cooling mode, a cold plate mode, and an immersion liquid cooling mode; A second determining unit is configured to determine a pre-executed refrigeration logic of each server according to a refrigeration mode of each server, and acquire real-time operation data of each device in each server in real time during refrigeration of each server, extract an executing refrigeration logic from the real-time operation data, and determine whether the executing refrigeration logic is correct according to a difference between the pre-executed refrigeration logic and the executing refrigeration logic.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the steps of the method in any one of claims 1 to 6.

9. A server refrigeration system characterized by, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the steps of the method in any one of claims 1 to 6. The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the steps of the method in any one of claims 1 to 6. The server cluster comprises at least one server, and the server cluster communicates with the baseboard management controller.

10. The server chiller system of claim 9, wherein, The server refrigeration system further comprises an external memory, and the baseboard management controller has a first interface, the baseboard management controller is connected with the external memory through the first interface, and the external memory is configured to store the refrigeration logic.

11. The server chiller system of claim 9, wherein, The baseboard management controller further has at least one second interface, and the baseboard management controller is connected with one server through one second interface.

Images