Aging rack distribution, flow control method and system in aging test

Abstract

The application discloses an aging rack distribution and flow control method and system in an aging test, relates to the technical field of heat dissipation, and comprises the following steps: obtaining a distribution request corresponding to an electronic device to be tested; analyzing the distribution request to obtain test parameters corresponding to the electronic device to be tested; calling real-time layout data of a pre-constructed aging rack resource database according to the cold plate attribute information, and screening at least one candidate aging rack position; and selecting a target aging rack position from the at least one candidate aging rack position according to the position constraint condition. By adopting the method, the risk of electrochemical corrosion caused by the distribution of cold plates with different materials under the same CDU can be avoided. Through accurate matching of the cold plate material and the interface, test interruption caused by physical incompatibility can also be avoided. The method does not rely on manual distribution of the aging rack position of the electronic device to be tested, and the error-prone problem of manual partition management is avoided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of heat dissipation, in particular to an aging rack distribution method and system and a flow control method in an aging test. BACKGROUND

[0002] With the exponential growth of data center computing density and power density, traditional air-cooled heat dissipation solutions have been difficult to meet the heat dissipation needs of high-performance electronic devices, especially artificial intelligence servers. Liquid cooling heat dissipation technology, with its excellent heat conduction efficiency, ultra-low noise characteristics, and more than 40% potential for optimizing power usage efficiency, has become the core technology route to break through the heat dissipation bottleneck. Among them, the cold plate type liquid cooling occupies the mainstream position of current commercial applications due to its advantages of direct contact high-efficiency heat conduction and non-intrusive transformation.

[0003] Before the liquid cooling equipment is shipped, it must undergo strict burn-in test. The burn-in test occupies a large amount of factory space, power, and cooling resources, and takes a long time. How to efficiently and reasonably distribute the burn-in test location directly affects the burn-in production capacity and equipment utilization. SUMMARY

[0004] The present application provides an aging rack distribution method and system and a flow control method in an aging test to at least solve the problem that the aging test location cannot be reasonably distributed in the related art.

[0005] The present application provides an aging rack distribution method, which comprises:

[0006] Obtaining a distribution request corresponding to an electronic device to be tested;

[0007] Analyzing the distribution request to obtain test parameters corresponding to the electronic device to be tested, the test parameters including cold plate attribute information and test location constraint conditions;

[0008] According to the cold plate attribute information, calling real-time layout data of a pre-constructed aging rack resource database to filter at least one candidate aging rack location;

[0009] According to the location constraint conditions, selecting a target aging rack location from the at least one candidate aging rack location to place the electronic device to be tested in the target aging rack location to complete the aging test of the electronic device to be tested.

[0010] The present application provides a flow control method in an aging test, which comprises:

[0011] Obtaining total target flow data corresponding to a target cooling distribution unit sent by a distribution engine, the total target flow data being flow data corresponding to a main pipeline in the target cooling distribution unit;

[0012] According to the total target flow data, the pre-acquired pressure difference at both ends of the valve of the pipeline in the target cooling distribution unit, and the specific gravity of the fluid, the valve opening degree of the pipeline in the target cooling distribution unit is determined, so as to adjust the valve flow of the pipeline to the total target flow data.

[0013] The application also provides an aging rack distribution device, which comprises:

[0014] An acquisition module is configured to acquire a distribution request corresponding to the electronic device to be tested.

[0015] An analysis module is configured to analyze the distribution request and acquire test parameters corresponding to the electronic device to be tested, wherein the test parameters comprise cold plate attribute information and test position constraint conditions.

[0016] A calling module is configured to call real-time layout data of a pre-constructed aging rack resource database according to the cold plate attribute information, and filter at least one candidate aging rack position.

[0017] A processing module is configured to select a target aging rack position from the at least one candidate aging rack position according to the position constraint conditions, and place the electronic device to be tested in the target aging rack position to complete the aging test of the electronic device to be tested.

[0018] The application also provides an aging test flow control device, which comprises:

[0019] An acquisition module is configured to acquire total target flow data corresponding to a target cooling distribution unit sent by a distribution engine, wherein the total target flow data is flow data corresponding to a main pipeline in the target cooling distribution unit.

[0020] An adjustment module is configured to determine a valve opening degree of a pipeline in the target cooling distribution unit according to the total target flow data, pre-acquired pressure difference at both ends of the valve of the pipeline in the target cooling distribution unit, and specific gravity of the fluid, so as to adjust valve flow of the pipeline to the total target flow data.

[0021] Embodiments of the application also provide an aging test system, which comprises a distribution engine and a cooling distribution unit controller.

[0022] The distribution engine is configured to perform steps of the aging rack distribution method as described above.

[0023] The cooling distribution unit controller is configured to perform steps of the aging test flow control method as described above.

[0024] The application also provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program is executed by the distribution engine to implement steps of the aging rack distribution method.

[0025] Alternatively, the computer program, when executed by the cooling distribution unit controller, implements the steps of the flow control method described above in the aging test.

[0026] This application also provides a computer program product, including a computer program that, when executed by an allocation engine, implements the steps of the aging rack allocation method described above.

[0027] Alternatively, the computer program, when executed by the cooling distribution unit controller, implements the steps of the flow control method described above in the aging test.

[0028] This application classifies the CDUs (Continuous Cooling Units) to which the electronic devices under test belong using cold plate attribute information. Real-time layout data of the CDU piping enables physical isolation of cold plate materials. This avoids the risk of electrochemical corrosion caused by assigning cold plates of different materials to the same CDU. Precise matching of cold plate materials and interfaces also prevents test interruptions (such as leaks or connection failures) due to physical incompatibility. This method does not rely on manual allocation of burn-in rack positions for the electronic devices under test; instead, it uses an automated isolation mechanism to avoid the errors that are prone to occur in manual partitioning management. The real-time database reduces manual troubleshooting time and improves burn-in rack occupancy. Multi-dimensional constraints are supported, reducing deployment failure rates.

[0029] Furthermore, because each CDU in this application corresponds to an electronic device with a cold plate attribute, compared to multiple CDUs in the same area that can only be used to store electronic devices with the same type of cold plate attribute, this reduces resource waste and avoids the occurrence of low utilization of aging test bits. Attached Figure Description

[0030] To more clearly illustrate the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 This is a schematic diagram of an aging rack allocation method provided in an embodiment of this application;

[0032] Figure 2 This is a schematic diagram of another aging rack allocation method provided in an embodiment of this application;

[0033] Figure 3 A schematic diagram of a CDU pipeline structure provided in an embodiment of this application;

[0034] Figure 4 This is a schematic diagram of another aging rack allocation method provided in an embodiment of this application;

[0035] Figure 5 Another aging rack allocation method flow chart provided by the embodiment of the present application;

[0036] Figure 6 A flow chart of a flow control method in an aging test provided by the embodiment of the present application;

[0037] Figure 7 An overall principle structure schematic diagram provided by the embodiment of the present application;

[0038] Figure 8 An overall operation timing flow chart provided by the embodiment of the present application;

[0039] Figure 9 An aging rack allocation device structure schematic diagram provided by the embodiment of the present application;

[0040] Figure 10 An aging test flow control device structure schematic diagram provided by the embodiment of the present application;

[0041] Figure 11 An aging test system structure schematic diagram provided by the embodiment of the present application. DETAILED DESCRIPTION

[0042] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without any creative work fall within the protection scope of the present application.

[0043] It should be noted that, in the description of the present application, the terms “include”, “contain” or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or includes elements inherent to such process, method, article or device. The terms “first”, “second” and the like in the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence.

[0044] In order for those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0045] With the exponential growth of computing density and power density in data centers, traditional air-cooling solutions have been difficult to meet the heat dissipation needs of certain electronic devices, such as high-performance electronic devices (artificial intelligence servers, switches, storage devices), etc. In the following, liquid-cooled servers will be taken as an example for illustration. In a typical AI training cluster, the single-cabinet power density generally breaks through 30kW, and the local hotspot temperature can reach above 90℃. Liquid cooling technology has become the core technology route to break through the heat dissipation bottleneck, with its outstanding heat conduction efficiency (3-5 times higher than air cooling), ultra-low noise characteristics below 35dB, and more than 40% potential for optimizing power usage effectiveness (PUE).

[0046] Before the liquid-cooled servers are shipped, they must undergo strict burn-in tests. This test needs to be run continuously for 48-72 hours under ≥90% load conditions, simulating the early running environment of the server to accelerate the exposure of potential material defects, virtual connection of welding, early failure of components, etc. According to statistics, effective burn-in tests can reduce the server's field failure rate by more than 60%, which is a key quality barrier to ensure long-term reliable operation of the product.

[0047] The current cold plate materials used in liquid-cooled servers mainly include copper and aluminum alloys. When using water-based coolant, if copper and aluminum, two different metals, exist in the same liquid cooling circuit, galvanic corrosion will occur. In the presence of conductive coolant, aluminum will accelerate dissolution and corrosion, leading to cold plate perforation, coolant leakage, and ultimately causing server damage and even safety accidents, which is a technical risk that must be strictly avoided in the burn-in test of liquid-cooled servers.

[0048] At the same time, due to differences in power, heat dissipation design, and cold plate type, different models and configurations of liquid-cooled servers have obvious differentiated needs for coolant type, flow rate, and pipe joint model, etc. Too high flow rate will cause energy waste and increase the burden on the cooling system; too low flow rate will not effectively remove heat, leading to high temperature errors during server testing, affecting test effectiveness, and even damaging equipment.

[0049] Burn-in tests occupy a large amount of factory space, power, and cooling resources, and take a long time. How to efficiently and reasonably allocate the burn-in tests according to the specific parameters of the servers to be tested (such as flow rate requirements, cold plate material, size, power consumption) and the current state of the burn-in site (location occupation, CDU load) directly affects the burn-in capacity and equipment utilization.

[0050] However, in the related art, the general aging rack position management is only used to record the occupation state of the aging rack to facilitate the administrator to find the idle position, and does not have the allocation position function, so that the copper and aluminum cold plates of different materials may be mixed in the same liquid cooling loop to cause galvanic corrosion; and the liquid cooling flow required by different models of servers is different, which may also cause energy waste or too low flow to effectively take away the heat.

[0051] To solve the copper and aluminum corrosion risk, the most common method is to artificially and fixedly mark the aging area or specific aging rack / CDU loop as a "copper area" or an "aluminum area". The operator needs to manually put the corresponding server into the specified area.

[0052] The fixed partition or simple rule isolation has the disadvantages of resource waste, poor flexibility, error prevention, etc. For example, if only copper cold plates need to be tested at present, the "aluminum area" resources are idle, causing low utilization rate of the aging test site, and vice versa. It is impossible to dynamically allocate according to the real-time copper and aluminum cold plate test demand, resulting in poor flexibility. Moreover, there is still a risk of mistakenly putting aluminum cold plates into the "copper area" by manual operation.

[0053] To solve the above problems, an embodiment of the present application provides an aging rack allocation method, which is specifically described with reference to Figure 1 as shown in the figure, Figure 1 The aging rack allocation method provided by the embodiment of the present application is a flowchart, and specifically includes the following method steps:

[0054] Step S101, an allocation request corresponding to the electronic device to be tested is obtained.

[0055] Step S102, the allocation request is parsed to obtain the test parameters corresponding to the electronic device to be tested.

[0056] Specifically, taking a server as an example, the scanning terminal first queries the server test parameters corresponding to the SN code through the MES interface, including the cold plate material (such as copper or aluminum), the cold plate interface type, and the test position constraint condition, etc. After obtaining the test parameters, the scanning terminal submits the allocation request to the intelligent allocation engine. The test parameters of the server to be tested will be included in the allocation request. Therefore, the intelligent allocation engine can obtain the corresponding test parameters after parsing the allocation request.

[0057] Step S103, according to the cold plate attribute information, the real-time layout data of the pre-constructed aging rack resource database is called to screen at least one candidate aging rack position for carrying the electronic device to be tested.

[0058] Specifically, in one specific example, the cold plate attribute information may include, for example, cold plate material and cold plate interface type. The cold plate material is, for example, the copper material, the aluminum material, and the like introduced above. In the cold plate isolation coarse screening stage of the intelligent allocation engine, the cold plate interface type usually includes the following key categories for matching with the CDU pipeline in the aging rack resource database:

[0059] 1. Physical interface standard

[0060] Quick Connect

[0061] For example: industrial standard quick connector interfaces such as CPC, Legris, Parker, etc., distinguishing between male / female, with or without self-locking.

[0062] Flange

[0063] Need to match the flange specification (such as DN size, bolt hole position), sealing surface type (flat / convex).

[0064] Threaded

[0065] Including NPT, BSPP, BSPT, etc. Threaded standards, need to distinguish size (such as 1 / 4", 3 / 8").

[0066] 2. Fluid property interface

[0067] Pipe diameter and flow rate

[0068] Interface inner diameter (such as Φ6mm, Φ8mm) and maximum flow compatibility (such as 10L / min).

[0069] Flow direction identification

[0070] One-way / two-way flow design, directional allocation requirements of inlet / outlet.

[0071] 3. Electrical / signal interface (if it is a smart cold plate)

[0072] Sensor interface

[0073] Temperature / flow sensor cable connector type (such as M12 aviation plug, RJ45).

[0074] Communication protocol

[0075] Support Modbus, CAN bus, etc. Protocol electrical interface.

[0076] 4. Sealing and pressure rating

[0077] Sealing type

[0078] O-ring seal (material needs to be compatible, such as fluorine rubber), metal seal, etc.

[0079] Pressure range

[0080] Low pressure (≤1MPa), medium pressure (1-3MPa), high pressure (≥3MPa) interface differentiation.

[0081] 5. Special design interface

[0082] All-in-one composite interface

[0083] Integrated fluid + electrical integrated interface (such as cooling + power hybrid plug).

[0084] Foolproof design

[0085] Special-shaped interface or color coding to avoid misplug (such as Keyed Connector).

[0086] According to the cold plate attribute information, the real-time layout data of the pre-constructed aging rack resource data can be called to determine the corresponding available CDU pipeline independent partition, and then the available aging rack in the CDU pipeline partition can be determined, and finally the specific available aging rack position can be determined.

[0087] Step S104, according to the position constraint condition, selecting a target aging rack position from at least one candidate aging rack position.

[0088] Specifically, different to-be-tested electronic devices are configured with different position constraint conditions, and then after obtaining at least one candidate aging rack position, in order to more accurately determine which specific position of the aging rack the to-be-tested electronic device is placed in, the specific position constraint condition needs to be matched for further determination.

[0089] The aging rack allocation method provided by the embodiment of the application classifies the CDU to which the to-be-tested electronic device belongs by using the cold plate attribute information. The real-time layout data of the CDU pipeline is used to realize physical isolation of the cold plate material. The risk of electrochemical corrosion caused by allocation of cold plates of different materials under the same CDU can be avoided. Through accurate matching of the cold plate material and the interface, test interruption (such as leakage or connection failure) caused by physical incompatibility can also be avoided. The method does not rely on manual allocation of the aging rack position of the to-be-tested electronic device, but uses an automatic isolation mechanism to avoid the error-prone problem of manual partition management. The real-time database reduces the manual troubleshooting time and improves the aging rack occupancy rate. Multi-dimensional constraints are supported to reduce the deployment failure rate.

[0090] In addition, because each CDU in the present application corresponds to an electronic device of a cold plate attribute, compared to multiple CDUs in the same area being used to store electronic devices of the same type of cold plate attribute, resource waste is reduced, and the situation of low utilization of burn-in test sites is avoided.

[0091] In an optional example, on the basis of the foregoing embodiment, according to the cold plate attribute information, real-time layout data of a pre-constructed burn-in rack resource database is called to screen at least one candidate burn-in rack position for carrying the electronic device to be tested, including the following method steps, which are specifically described with reference to Figure 2 as shown, including:

[0092] Step S201: According to the cold plate attribute information, a candidate cooling distribution unit that meets the cold plate attribute information is screened from real-time layout data of a burn-in rack resource database.

[0093] Step S202: At least one candidate burn-in rack position is screened according to the physical position relationship between the burn-in racks in the pre-acquired candidate cooling distribution unit.

[0094] Specifically, the real-time layout data at least contains the burn-in rack code corresponding to the CDU, the current cold plate material allocated by the CDU, the state information of each position of the burn-in rack (including whether there is an electronic device, the SN of the electronic device), etc. First, the same CDU as the current cold plate material allocation is found from the real-time layout information, the corresponding available burn-in rack is found according to the CDU, and then at least one specific candidate burn-in rack position that can be used is found.

[0095] The process of specifically determining the candidate burn-in rack position can be determined according to the physical connection relationship in the CDU pipeline.

[0096] Specifically, as shown in Figure 3 as shown, Figure 3 the CDU pipeline diagram is shown in Figure 3 The CDU pipeline diagram is shown in

[0097] Because Figure 3 The physical connection relationship shown in

[0098] In this method step, the compatibility problem caused by the test failure (such as leakage or corrosion) can be avoided through the double check of the material and the interface. The screening based on the topological relationship reduces the deployment path length, improves the efficiency of the automatic equipment, and shortens the test preparation time. The real-time data distribution candidate position is combined to avoid local overheating or overload. The method can also automatically exclude the aging rack with high interface wear rate to reduce the risk of connection failure.

[0099] On the basis of any of the foregoing embodiments, the position constraint condition comprises historical power consumption of the electronic device to be tested in at least one test item respectively. According to the position constraint condition, a target aging rack position is selected from at least one candidate aging rack position to complete the aging rack of the electronic device to be tested, and the method comprises the following method steps, which are specifically described with reference to Figure 4

[0100] In step S401, the power consumption attribute data corresponding to each candidate aging rack position and the total power data and total current data of the power distribution unit corresponding to the cooling distribution unit to which the candidate aging rack position belongs are obtained.

[0101] The power consumption attribute data comprises the target phase power line in the three-phase power line occupied by each candidate aging rack position, and the port voltage data and current data of the electronic device included in the candidate cooling distribution unit to which each candidate aging rack position belongs when tested in at least one test item.

[0102] In step S402, according to the target phase power line in the three-phase power line occupied by each candidate aging rack position, the port voltage data and current data of the electronic device included in the candidate cooling distribution unit to which each candidate aging rack position belongs when tested in at least one test item, the historical power consumption of the electronic device to be tested in at least one test item respectively, the total power data, and the total current data, a target aging rack position is selected from at least one candidate aging rack position.

[0103] Specifically, first, the power consumption of each electronic device placed on each aging rack in the candidate CDU is monitored in real time, for example, including the target phase power line in the three-phase power line occupied by each electronic device, the port voltage data and current data of each electronic device when tested in the preconfigured at least one test item, and the power consumption of the electronic device to be tested, for example, the historical power consumption of the electronic device to be tested in at least one test item respectively. At the same time, the total power data and total current data of the power distribution unit corresponding to the candidate cooling distribution unit to which the candidate aging rack position belongs also need to be determined.

[0104] ​Wherein, the specific phase of three-phase electricity corresponding to each aging bit has been known in the prior art, and the allocation process can ensure that the machine power corresponding to the three-phase electricity (U / V / W) is consistent. Different test items refer to the fact that the electronic device will perform some targeted tests on the components during the aging test process, such as CPU floating-point operation, memory stress test, hard disk read-write test, IO read-write test, and the like.

[0105] When a certain test item is actually performed, it is also necessary to consider whether the rated power of the CDU will be exceeded, and therefore it is necessary to know the total voltage data and total current data of the power distribution unit (PDU) corresponding to each candidate CDU, so as to determine the power consumption of all electronic devices that have been configured. The voltage data and current data of each electronic device port collected in the foregoing can reflect the power consumption of each electronic device at each test stage. Based on these data, when allocating the aging test position, it is necessary to avoid positions that may cause high power or current impact when starting a certain test item. Thus, more suitable positions for carrying the electronic device to be tested are selected. Moreover, high-power electronic devices can be distributed to different power phase lines to achieve phase balance and reduce the risk of power grid disturbance. Moreover, this method can also avoid the uneven distribution of power load in the machine room caused by the placement of all electronic devices of the same type in sequence, such as the case where a PDU in a certain area bears an excessive load, while the PDUs in other areas may be in a light load state. It can also avoid the accumulation of a large amount of heat generated during the aging test process of the electronic device, thereby reducing the risk of accelerated aging of the power supply device due to high temperature, leading to a decline in electrical performance, and accelerating the risk of local power overload.

[0106] When the determination is successful, the allocation engine feeds back the optimal rack number and port number to the scanning red satin, and simultaneously generates a set flow instruction.

[0107] In an optional example, in order to avoid the resource waste and increase labor costs caused by the fact that the flow of the CDU is roughly estimated by the operator according to experience or based on the full load of the aging rack, or directly set a relatively high fixed value, the embodiment of the present application further includes the following method steps, which are specifically described with reference to Figure 5 as shown, specifically comprising:

[0108] Step S501: acquiring the current set flow of each branch pipe in the target cooling distribution unit to which the target aging rack position belongs, and the temperature difference between the inlet and outlet flows in the target cooling distribution unit.

[0109] Step S502: determining the total target flow data of the target cooling distribution unit according to the current set flow and the temperature difference.

[0110] Specifically, in an optional example, according to the current set flow, the temperature difference value, the total target flow data of the target cooling distribution unit is determined, see the following expression:

[0111]

[0112] Wherein, is the total target flow data, is the current set flow of each branch pipe, is the material correction coefficient, is the temperature difference value at the current moment, is the refrigerant flow efficiency coefficient, i is the i-th electronic device in the target aging rack position, n is the number of all electronic devices carried in the target aging rack position, and each electronic device corresponds to a branch pipe.

[0113] Specifically, each branch pipe corresponds to an electronic device, and the sum of the flow data of all branch pipes is the total target flow data of the CDU main pipe.

[0114] In a specific example, it is assumed that three servers A, B and C are connected under the CDU main pipe, the server cold plate material is copper, the current set flow is 1000 L / min, is the material correction coefficient, which is 0.01 / ℃, is 95%.

[0115] When only server A is accessed, is 5℃, and is calculated as:

[0116]

[0117] When only server B is accessed, is 10℃, and is calculated as:

[0118]

[0119] When only server C is accessed, is 25℃, and is calculated as:

[0120] .

[0121] Of course, if all servers in the main pipe are accessed, then the sum of the above three final results.

[0122] Step S503: Send the total target flow data to the cooling distribution unit controller corresponding to the target cooling distribution unit.

[0123] Specifically, the purpose of sending the total target flow rate data to the cooling distribution unit controller corresponding to the target cooling distribution unit is to enable the cooling distribution unit controller to determine the valve opening degree of the pipeline of the target cooling distribution unit based on the total target flow rate data. The valve opening degree is used to control the flow rate of refrigerant in the cooling distribution unit pipeline.

[0124] This method allows for real-time and precise flow control based on the dynamically changing heat dissipation requirements of electronic devices within the CDU circuit, avoiding energy waste or insufficient heat dissipation risks caused by setting target flow rates based on human experience or other methods in related technologies.

[0125] In an optional embodiment, the method further includes the following method steps:

[0126] Step a1: Obtain the ambient temperature in real time;

[0127] Step a2: When it is determined that the difference in ambient temperature between two adjacent moments is greater than the preset temperature threshold, the compensation flow is determined according to the preset flow compensation algorithm, and the flow compensation operation is triggered.

[0128] Specifically, considering that increased ambient temperature may weaken coolant heat exchange, it is necessary to increase the flow rate to maintain the total heat transfer. Therefore, the method also includes determining the compensation flow rate according to a preset flow compensation algorithm and triggering the flow compensation operation.

[0129] In a specific example, the expression for the preset traffic compensation algorithm is shown below:

[0130]

[0131] in, To compensate for the data usage, The cold plate coefficient corresponds to the cold plate material included in the cold plate attribute information. This represents the highest test temperature corresponding to the electronic device under test. This represents the lowest test temperature corresponding to the electronic device under test. This represents the difference in ambient temperature.

[0132] This application also provides a flow control method for aging tests, see details below. Figure 6 As shown, Figure 6 This application provides a schematic flowchart of a flow control method for aging tests, which includes:

[0133] Step S601, obtaining total target flow data corresponding to the target cooling distribution unit sent by the distribution engine.

[0134] Specifically, as introduced before, the total target flow data is flow data corresponding to the main pipe in the target cooling distribution unit, which is the refrigerant flow data. The target cooling distribution unit is the target cooling distribution unit mentioned in the foregoing embodiments.

[0135] Step S602, determining the valve opening degree of the pipe in the target cooling distribution unit according to the total target flow data, the pre-obtained pressure difference between the two ends of the valve of the pipe in the target cooling distribution unit, and the specific gravity of the fluid, so as to adjust the valve flow of the pipe to the total target flow data.

[0136] In an optional example, the valve opening degree of the pipe in the target cooling distribution unit is determined according to the total target flow data, the pre-obtained pressure difference between the two ends of the valve of the pipe in the target cooling distribution unit, and the specific gravity of the fluid, and is specifically expressed by the following expression:

[0137]

[0138] wherein, is the valve opening degree, is the current set flow of each branch pipe in the target cooling distribution unit, is the pressure difference between the two ends of the valve, and G is the specific gravity of the fluid, is the inverse function of the valve flow coefficient function.

[0139] wherein, is the fluid adjustment factor, and the pressure difference is greater, the required opening degree is smaller, the high pressure difference fluid is easier to pass, the specific gravity G is greater, the liquid is more viscous, and the opening degree needs to be increased. is the characteristic curve of the valve, that is, the flow capacity of different valves is different under the same opening degree. The inverse function will calculate the opening degree position of the flow according to the flow coefficient curve of the valve factory calibration, so as to realize the condition of zero corrosion of the cold plate under the same CDU pipe, that is, only one kind of material of the cold plate exists, so that the corrosion does not occur.

[0140] In a specific example, it is assumed that the flow set is 1000 L / min, the valve pressure difference is 5 psi, and the fluid specific gravity is 0.98 (wherein the specific gravity of water at a temperature of 60°F (15.6°C) is 1);

[0141] In the above formula for calculating the valve opening degree, first, the unit conversion of needs to be performed, and specific reference can be made to the following:

[0142]

[0143]

[0144] wherein, is the valve flow coefficient function, defined as the number of US gallons of water, at 60°F (15.6°C), that will flow through a valve per minute to produce a 1 psi pressure drop (ΔP) across the valve, with units defined as 1 .

[0145] Then all the known data obtained above are brought into , and the characteristics of and opening are queried to obtain = 25°.

[0146] Figure 7 The overall principle structure of the embodiment of the present application is shown in Figure 6 , which includes a scanning terminal, an MES interface, an intelligent distribution engine, a power load detection module, an intelligent PDU, an aging rack resource database, a CDU control interface, and an adjustable flow valve group. The method steps performed by each structure have been described in detail in the foregoing, and thus will not be described in detail here.

[0147] Figure 8 The overall operation timing flowchart of the method is shown, and the electronic device is taken as an example of a server, for example, including that an operator scans a server serial number (SN) code through a scanning terminal, the scanning terminal requests test parameters from a manufacturing execution system (MES), the MES system feeds back test parameters of the server to be tested, such as cold plate material, flow, power, and the like. Then, the terminal submits a distribution request to the distribution engine. The distribution engine performs a cold plate algorithm, calculates power phase balance, and the like, and when it is determined that there is a feasible solution, feeds back a rack ID and a port number to the scanning terminal. At the same time, a target flow is set, and is sent to a CDU controller. The CDU controller determines a valve opening, and adjusts the valve opening through an adjustable valve group.

[0148] When there is no available resource, the distribution engine also needs to send an alarm prompt message to the scanning terminal.

[0149] The above processes have been described in detail in the foregoing, and thus will not be described in detail here.

[0150] Compared with the traditional cooling system relying on fixed valve opening or manual adjustment, it is difficult to adapt to dynamic load changes. The above method of the embodiment of the application can ensure that the flow is accurately matched with the target demand by calculating the valve opening in real time, thereby avoiding excess or insufficient flow, and improving the system energy efficiency. Moreover, in the process, the pressure difference and fluid characteristics are considered comprehensively, the calculation is more accurate, and the flow deviation is reduced. The pressure difference parameter participates in the calculation, which can offset the influence of pressure fluctuation. Real-time feedback of flow data can dynamically correct the valve opening. Therefore, accurate heat dissipation can prolong the service life of the equipment, and can be applied to different fluids and pressure environments.

[0151] Through the description of the above embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be realized by means of software and the necessary general hardware platform, of course, it can also be realized by hardware, but in many cases the former is a better embodiment.

[0152] The embodiment of the application further provides an aging rack distribution device, please refer to Figure 9 The device is applied to a distribution engine, and the device comprises an acquisition module 901, an analysis module 902, a calling module 903, and a processing module 904.

[0153] The acquisition module 901 is used for acquiring a distribution request corresponding to a to-be-tested electronic device.

[0154] The analysis module 902 is used for analyzing the distribution request, acquiring test parameters corresponding to the to-be-tested electronic device, and the test parameters comprising cold plate attribute information and test position constraint conditions.

[0155] The calling module 903 is used for calling real-time layout data of a pre-constructed aging rack resource database according to the cold plate attribute information, and screening at least one candidate aging rack position.

[0156] The processing module 904 is used for selecting a target aging rack position from the at least one candidate aging rack position according to the position constraint conditions, placing the to-be-tested electronic device at the target aging rack position, and completing the aging test of the to-be-tested electronic device.

[0157] In an optional example, the calling module 903 is specifically used for screening a candidate cooling distribution unit conforming to the cold plate attribute information from the real-time layout data of the aging rack resource database according to the cold plate attribute information.

[0158] At least one candidate aging rack position is screened according to the physical position relationship between the aging racks in the pre-acquired candidate cooling distribution unit.

[0159] In an optional example, the position constraint includes historical power consumption of the electronic device to be tested in the at least one test item respectively; the processing module 904 is specifically configured to:

[0160] obtain power consumption attribute data corresponding to each candidate aging rack position, and total power data and total current data of the power distribution unit corresponding to the candidate cooling distribution unit to which the candidate aging rack position belongs, wherein the power consumption attribute data includes a target phase power line in a three-phase power line occupied by each candidate aging rack position, and each port voltage data and current data of the electronic device included in the candidate cooling distribution unit to which each candidate aging rack position belongs when tested in the at least one test item;

[0161] According to the target phase power line in the three-phase power line occupied by each candidate aging rack position, the each port voltage data and current data of the electronic device included in the candidate cooling distribution unit to which each candidate aging rack position belongs when tested in the at least one test item, the historical power consumption of the electronic device to be tested in the at least one test item respectively, the total power data, and the total current data, the target aging rack position is selected from the at least one candidate aging rack position.

[0162] In an optional example, the apparatus further includes a sending module 905;

[0163] The obtaining module 901 is further configured to obtain a current set flow of each branch pipe in the target cooling distribution unit to which the target aging rack position belongs, and a temperature difference between inlet and outlet flows in the target cooling distribution unit;

[0164] The processing module 904 is further configured to determine total target flow data of the target cooling distribution unit according to the current set flow and the temperature difference;

[0165] The sending module 905 is configured to send the total target flow data to a cooling distribution unit controller corresponding to the target cooling distribution unit, so that the cooling distribution unit controller determines a valve opening degree of a pipe of the target cooling distribution unit according to the total target flow data.

[0166] In an optional example, the processing module 904 determines the total target flow data of the target cooling distribution unit according to the current set flow and the temperature difference, and specifically refers to the following expression:

[0167]

[0168] wherein, the total target flow data is, the current set flow of each branch pipe is, the material correction coefficient is, the temperature difference at the current moment is, is a refrigerant flow efficiency coefficient, i is an i-th electronic device in a target aging rack position, n is a number of all electronic devices carried in the target aging rack position, and each electronic device corresponds to a branch pipeline.

[0169] In an optional example, the acquisition module 901 is further configured to acquire the environment temperature in real time.

[0170] The processing module 904 is further configured to determine a compensation flow according to a preset flow compensation algorithm and trigger a flow compensation operation when it is determined that the environment temperature change difference between two adjacent time points is greater than a preset temperature threshold.

[0171] In an optional example, the preset flow compensation algorithm expression is as follows:

[0172]

[0173] wherein, is a compensation flow, is a cold plate coefficient corresponding to a cold plate material included in the cold plate attribute information, is a highest test temperature corresponding to the electronic device to be tested, is a lowest test temperature corresponding to the electronic device to be tested, is the environment temperature change difference.

[0174] The features of the embodiments of the aging rack allocation device provided by the embodiments of the present application can be referred to the related descriptions of the embodiments of the aging rack allocation method, which will not be repeated here.

[0175] The aging rack allocation device provided by the embodiments of the present application classifies the CDU to which the electronic device to be tested belongs by using the cold plate attribute information. The physical isolation of the cold plate material is realized by using the real-time layout data of the CDU pipeline. The risk of electrochemical corrosion caused by the allocation of cold plates of different materials to the same CDU can be avoided. Through the accurate matching of the cold plate material and the interface, the test interruption (such as leakage or connection failure) caused by physical incompatibility can also be avoided. This way does not rely on manual allocation of the aging rack position of the electronic device to be tested, but uses an automatic isolation mechanism to avoid the error-prone problem of manual partition management. The real-time database reduces the manual troubleshooting time and improves the aging rack occupancy rate. It supports multi-dimensional constraints and reduces the deployment failure rate.

[0176] In addition, because each CDU in the present application corresponds to an electronic device of a cold plate attribute, compared to the fact that multiple CDUs in the same area can only be used to store electronic devices of the same type of cold plate attribute, resource waste is reduced and the situation of low utilization rate of aging test sites is avoided.

[0177] The embodiment of the present application also provides a flow control device in an aging test, which can be specifically seen from Figure 10 As shown in the figure, the device is applied to a cooling distribution unit controller, and the device comprises an acquisition module 1001 and an adjustment module 1002.

[0178] The acquisition module 1001 is used for acquiring total target flow data corresponding to a target cooling distribution unit sent by a distribution engine, wherein the total target flow data is flow data corresponding to a main pipeline in the target cooling distribution unit;

[0179] The adjustment module 1002 is used for determining a valve opening degree of a pipeline in the target cooling distribution unit according to the total target flow data, a pre-acquired pressure difference between two ends of a valve of the pipeline in the target cooling distribution unit and a specific gravity of a fluid, so as to adjust a valve flow of the pipeline to be the total target flow data.

[0180] In an optional example, the adjustment module 1002 determines the valve opening degree of the pipeline in the target cooling distribution unit according to the total target flow data, the pre-acquired pressure difference between the two ends of the valve of the pipeline in the target cooling distribution unit and the specific gravity of the fluid, and specifically adopts the following expression to represent:

[0181]

[0182] wherein, is the valve opening degree, is a current set flow of each branch pipeline in the target cooling distribution unit, is the pressure difference between the two ends of the valve, and G is the specific gravity of the fluid, is an inverse function of a valve flow coefficient function.

[0183] The description of the features in the embodiment of the flow control device in the aging test provided by the present application can be referred to the related description of the embodiment of the flow control method in the aging test, which will not be repeated here.

[0184] Compared with the traditional cooling system which relies on fixed valve opening degrees or manual adjustment and is difficult to adapt to dynamic load changes, the flow control device in the aging test provided by the embodiment of the present application can ensure that the flow is accurately matched with the target demand by calculating the valve opening degree in real time, so as to avoid excess or insufficient flow and improve the system energy efficiency. Moreover, in the process, the pressure difference and the fluid characteristics are comprehensively considered, the calculation is more accurate, and the flow deviation is reduced. The pressure difference parameter participates in the calculation, which can offset the influence of pressure fluctuation. And the real-time feedback flow data can dynamically correct the valve opening degree. Therefore, the heat dissipation is accurate, the equipment life is prolonged, and the device can be applied to different fluid and pressure environments.

[0185] The embodiment of the present application also provides an aging test system, which can be specifically seen from Figure 11As shown, the system comprises: an allocation engine 1101 and a cooling allocation unit controller 1102;

[0186] The allocation engine 1101 is configured to perform the steps of the burn-in rack allocation method as described in any of the preceding embodiments.

[0187] The cooling allocation unit controller 1102 is configured to perform the steps of the flow control method in burn-in test as described in any of the preceding embodiments.

[0188] Embodiments of the present application also provide a computer readable storage medium, which stores a computer program, wherein the computer program is configured to perform the steps of any of the burn-in rack allocation method embodiments or the steps of any of the flow control method in burn-in test embodiments when executed.

[0189] In an example embodiment, the computer readable storage medium can include, but is not limited to, a U disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store computer programs.

[0190] Embodiments of the present application also provide a computer program product, which comprises a computer program, and the computer program is executed by a processor to implement the steps of any of the burn-in rack allocation method embodiments.

[0191] Embodiments of the present application also provide another computer program product, which comprises a non-volatile computer readable storage medium, and the non-volatile computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of any of the flow control method in burn-in test embodiments.

[0192] Those skilled in the art will further appreciate that the units and algorithm steps of the examples described in connection with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, various components and steps have been described above generally in terms of their functionality, without limitation. The specific implementation of the described functionality no matter whether it is implemented in hardware or in software depends on the particular application and design constraints imposed on the overall system. Those skilled in the art can implement the described functionality in varying ways for each particular application, but such implementation should not be construed to limit the scope of the present application.

[0193] The above describes in detail the method and system for flow control in the aging rack distribution and aging test provided by the present application. The principles and implementation modes of the present application are described by using specific examples in this paper, and the above description of the examples is only applicable to help understand the method of the present application and its core idea. It should be pointed out that, for ordinary skilled persons in the technical field, some improvements and modifications can be made to the present application without departing from the principles of the present application, and these improvements and modifications also fall within the protection scope of the claims of the present application.

Claims

1. A method for distributing aging racks, characterized in that, The method includes: Obtain the allocation request corresponding to the electronic device under test; The allocation request is parsed to obtain the test parameters corresponding to the electronic device under test. The test parameters include cold plate attribute information and position constraints of the test position. Based on the cold plate attribute information, candidate cooling allocation units that match the cold plate attribute information are selected from the real-time layout data of the aging rack resource database. The real-time layout data includes at least: the aging rack code corresponding to each cooling allocation unit, the currently allocated cold plate material, and the status information of each position of the aging rack. Based on the pre-obtained physical positional relationship between each aging rack in the candidate cooling distribution unit, at least one candidate aging rack position is selected; According to the position constraints, a target aging rack position is selected from at least one candidate aging rack position to place the electronic device under test in the target aging rack position to complete the aging test of the electronic device under test. The method further includes: Obtain the current set flow rate of each branch pipe in the target cooling distribution unit to which the target aging rack is located, and the temperature difference between the inlet and outlet flow rates in the target cooling distribution unit; Based on the current set flow rate and the temperature difference, determine the total target flow rate data of the target cooling distribution unit; The total target flow rate data is sent to the cooling distribution unit controller corresponding to the target cooling distribution unit, so that the cooling distribution unit controller can determine the valve opening of the pipeline of the target cooling distribution unit based on the total target flow rate data; Specifically, the determination of the total target flow rate data of the target cooling distribution unit based on the current set flow rate and the temperature difference is shown in the following expression: in, The total target traffic data, Set the current flow rate for each branch pipe. This is the material correction factor. The temperature difference value at the current moment. Let i be the refrigerant flow efficiency coefficient, i be the i-th electronic device in the target aging rack position, n be the total number of electronic devices carried in the target aging rack position, and each electronic device corresponds to one branch pipe.

2. The method according to claim 1, characterized in that, The location constraints include the historical power consumption of the electronic device under test in at least one test item; the step of selecting a target aging rack position from at least one candidate aging rack position according to the location constraints to complete the aging rack of the electronic device under test specifically includes: Obtain the power consumption attribute data corresponding to each of the candidate aging rack locations, as well as the total power data and total current data of the power distribution unit corresponding to the candidate cooling distribution unit to which the candidate aging rack location belongs. The power consumption attribute data includes the target phase power line in the occupied three-phase power line, the port voltage data and current data of the electronic equipment included in the candidate cooling distribution unit to which each candidate aging rack location belongs when tested in at least one of the test items. The target aging rack position is selected from at least one candidate aging rack position based on the target phase power line in the three-phase power line occupied by each candidate aging rack position, the port voltage data and current data of the electronic equipment included in the candidate cooling distribution unit to which each candidate aging rack position belongs during testing in at least one of the test items, the historical power consumption of the electronic equipment under test in at least one test item, the total power data, and the total current data.

3. The method according to claim 1, characterized in that, The method further includes: Real-time acquisition of ambient temperature; When the difference in ambient temperature between two adjacent moments is determined to be greater than a preset temperature threshold, the compensation flow is determined according to the preset flow compensation algorithm, and the flow compensation operation is triggered.

4. The method according to claim 3, characterized in that, The expression for the preset traffic compensation algorithm is as follows: in, For the compensation flow, This refers to the cold plate coefficient, which corresponds to the cold plate material included in the cold plate attribute information. This refers to the highest test temperature corresponding to the electronic device under test. This is the lowest test temperature corresponding to the electronic device under test. This refers to the difference in ambient temperature.

5. An aging test system, characterized in that, The aging test system includes: a distribution engine and a cooling distribution unit controller; The allocation engine is used to perform the steps of the aging rack allocation method according to any one of claims 1-4.

Images