Server system management methods and server systems
By real-time monitoring and weighted calculation of the physicochemical properties of the liquid working fluid, combined with the failure rate of the server and liquid cooling circulation system, early warning information and maintenance strategies are generated. This solves the problems of high failure rate and high maintenance cost in immersion liquid-cooled server systems, and realizes dynamic management of the system and reduces failure risk.
Patent Information
- Application Number
- CN202611340834.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-31
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2046-08-31
AI Technical Summary
Existing immersion liquid-cooled server systems lack real-time detection of the physicochemical properties of the liquid working fluid and early warning of failure risks, resulting in high system failure rates, high maintenance costs, and limited service life.
By acquiring multiple physicochemical indicators of the liquid working fluid in real time, calculating and weighting the independent failure rates, and combining them with the failure rates of the server and the liquid cooling circulation system, early warning information and maintenance strategies are generated to achieve dynamic management of the server system.
It enables dynamic prediction of the overall failure risk of immersion liquid-cooled server systems, reduces the probability of unplanned downtime, extends equipment lifespan, and reduces maintenance costs.
Smart Images

Figure CN122862070B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of artificial intelligence technology, and in particular to a server system management method and a server system. Background Technology
[0002] In immersion liquid-cooled server systems, the physicochemical properties of the liquid working fluid directly impact the reliability and safety of system operation. Existing server systems lack real-time monitoring of the liquid working fluid's physicochemical properties, dynamic calculation of the overall system failure rate, and early warning of failure risks. Repairs or replacements are only performed after a server malfunctions, resulting in high system failure rates, high maintenance costs, and limited lifespan. Summary of the Invention
[0003] This application provides a server system management method and a server system, which at least solves the problem in related technologies that repairs or replacements can only be carried out after a server malfunctions, resulting in high system failure rates, high maintenance costs, and limited service life.
[0004] This application provides a method for managing a server system, including: The system acquires the values of multiple physicochemical indicators of the liquid working fluid in the server system in real time, and calculates the independent failure rate of each physicochemical indicator based on the indicator values. The independent failure rate is the failure rate of the liquid working fluid caused by the corresponding physicochemical indicator. Determine the correlation coefficients among multiple physicochemical indicators, and calculate the independent failure rate by weighting the independent failure rate based on the correlation coefficients and the preset temperature influence coefficients corresponding to each of the multiple physicochemical indicators, so as to obtain the first failure rate of the liquid working fluid. Obtain the second failure rate of the server and the third failure rate of the liquid cooling circulation system in the server system. Add the second failure rate, the third failure rate and the first failure rate to obtain the target failure rate of the server system. In response to a target failure rate exceeding a preset failure rate threshold, a warning level is determined based on the difference between the target failure rate and the preset failure rate threshold, and corresponding fault warning information and maintenance strategies are generated based on the warning level.
[0005] This application also provides a server system, including: The detection system is used to detect the values of multiple physicochemical indicators of the liquid working fluid in the server system in real time. An automated maintenance system is used to perform maintenance on liquid working fluids. The central control system is connected to both the detection system and the automatic maintenance system, and is used to implement the management methods of the aforementioned server system.
[0006] This application also provides an electronic device, including: a memory for storing a computer program; and a processor for executing the computer program to implement the management method of any of the above-described server systems.
[0007] This application also provides a computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the steps of any of the above-described server system management methods.
[0008] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of any of the above-described server system management methods.
[0009] The server system management method and server system of this application can determine the degree of degradation of the liquid working fluid by acquiring the index values of multiple physicochemical indicators of the liquid working fluid in real time and calculating the independent failure rate of each physicochemical indicator. By using the correlation coefficient and temperature influence coefficient between physicochemical indicators to calculate the independent failure rate, the interaction between physicochemical indicators and the difference in the effect of temperature on failure can be accurately reflected. The server failure rate, the liquid cooling circulation system failure rate and the liquid working fluid failure rate are added together to obtain the system's target failure rate, which can realize the overall control of the server system. When the target failure rate exceeds the threshold, a graded warning is given according to the degree of exceedance and a corresponding maintenance strategy is generated. Maintenance is triggered and differentiated according to risk level, which can realize reasonable proactive maintenance, reduce the probability of downtime and maintenance costs, and extend the reliable operating life of the system. Attached Figure Description
[0010] 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.
[0011] Figure 1 A flowchart illustrating a server system management method provided in an embodiment of this application; Figure 2 This application provides a schematic diagram of an early warning process for a server system. Figure 3 This application provides a schematic diagram of an automated maintenance process. Figure 4 This is a schematic diagram of the structure of a server system provided in an embodiment of this application. Detailed Implementation
[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of this application.
[0013] It should be noted that, in the description of this application, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. The terms "first," "second," etc., in this application are used to distinguish similar objects and are not used to describe a specific order or sequence.
[0014] To enable 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 accompanying drawings and specific embodiments.
[0015] This application discloses a management method for a server system. The method is applied to an immersion liquid-cooled server system and aims to assess the health status of the system and provide graded early warning and maintenance strategies based on the assessment results, thereby effectively improving the security and reliability of the server system operation.
[0016] The server system of this invention refers to a server system employing immersion liquid cooling technology. This system mainly includes: the server itself for data processing, a housing for containing the server and storing liquid working fluid, a cooling distribution unit for driving the circulation of the liquid working fluid and achieving heat exchange, and liquid working fluid filled within the housing for directly absorbing the heat generated during server operation. The liquid working fluid is the medium for heat dissipation in the system, and the stability of its physicochemical properties directly affects the server's operational safety and heat dissipation efficiency.
[0017] Figure 1 This document provides a flowchart illustrating a server system management method as an embodiment of the present application. The method is described in detail below in conjunction with the execution flow of the server system management method.
[0018] Step 101: In real time, obtain the index values of multiple physicochemical indicators of the liquid working fluid in the server system, and calculate the independent failure rate of each of the multiple physicochemical indicators based on the index values. The independent failure rate is the failure rate of the liquid working fluid caused by the corresponding physicochemical indicator.
[0019] In the embodiments of this application, real-time acquisition refers to continuously or periodically collecting data through various sensors deployed in the system at a preset sampling frequency to ensure the timeliness of the system status.
[0020] Multiple physicochemical indicators are a set of parameters characterizing the physical and chemical properties of liquid working fluids, including but not limited to: kinematic viscosity characterizing the flow resistance and heat dissipation efficiency of liquid working fluids; dielectric constant, volume resistivity and breakdown voltage characterizing the insulation performance of liquid working fluids; pH value characterizing the acidity and alkalinity of liquid working fluids; and thermal conductivity and specific heat capacity characterizing the heat exchange capacity of liquid working fluids.
[0021] Independent failure rate is the probability or frequency of a liquid working fluid losing all or part of its specified function (i.e., failure) under the assumption that only a single physicochemical index deteriorates and ignores the interaction of other indices.
[0022] The independent failure rate can be calculated based on a pre-constructed mapping model between the physicochemical index value and the failure probability of the liquid working fluid (e.g., a mapping relationship fitted by historical experimental data).
[0023] Step 102: Determine the correlation coefficients between multiple physicochemical indicators, and calculate the independent failure rate by weighting the independent failure rate based on the correlation coefficients and the preset temperature influence coefficients corresponding to each of the multiple physicochemical indicators, so as to obtain the first failure rate of the liquid working fluid.
[0024] In the embodiments of this application, the correlation coefficient is an indicator used to measure the degree of linear correlation between two or more variables. In this application, it refers to the interrelationship between the changes in the values of different physicochemical indicators. For example, when water is mixed into a liquid working fluid, its dielectric constant will increase significantly, while its breakdown voltage will decrease sharply; these two indicators show a strong negative correlation. This correlation coefficient can be obtained by calculating based on historical datasets accumulated during system operation using the Pearson correlation coefficient or other suitable statistical methods.
[0025] The preset temperature influence coefficient is a weighting factor used to correct the degree of influence of temperature on various physicochemical properties. Since the physicochemical properties (such as kinematic viscosity) of liquid working fluids change with temperature, directly treating the values measured at different temperatures as equivalent will introduce errors. Therefore, the temperature influence coefficient allows the measured values of all physicochemical properties to be corrected to a unified standard temperature reference.
[0026] The specific implementation of the weighted calculation is as follows: For any first physicochemical indicator, the weight of the independent failure rate of the first physicochemical indicator is adjusted based on its correlation coefficient with one or more other second physicochemical indicators, as well as the independent failure rate of the second physicochemical indicators. Next, the adjusted independent failure rate is corrected using the temperature influence coefficient corresponding to each physicochemical indicator. Finally, the failure rates of all physicochemical indicators after double correction are summed or multiplied, and the result is the first failure rate.
[0027] The first failure rate represents the real-time failure probability of the liquid working fluid as a whole, after considering the inherent correlation between indicators and the influence of ambient temperature.
[0028] Step 103: Obtain the second failure rate of the server and the third failure rate of the liquid cooling circulation system in the server system. Add the second failure rate, the third failure rate and the first failure rate together to obtain the target failure rate of the server system.
[0029] In the embodiments of this application, the second failure rate is the failure rate of the server equipment itself in the server system. The server equipment includes electronic components such as the motherboard, central processing unit, memory, hard drive, and various interface cards. The failure rate of electronic equipment typically follows an exponential distribution, and its failure rate can be approximated as a constant over the equipment's lifespan. Therefore, the second failure rate can be predetermined as a constant value based on the server equipment's historical operating data or reliability indicators provided by the manufacturer. In one possible implementation, the second failure rate can also be dynamically adjusted based on factors such as the actual operating time and load rate of the server equipment.
[0030] The third failure rate is that of the liquid cooling circulation system in the server system. A liquid cooling circulation system refers to the assembly of equipment used to achieve the circulation of liquid working fluid and heat exchange with the external environment. It includes at least a cooling capacity distribution unit and an immersion enclosure, as well as key components such as circulation pipes, valves, seals, and water pumps connecting the two. During long-term operation, various key components of the liquid cooling circulation system will gradually lose their original performance due to wear, aging, fatigue, etc., leading to system failure. For example, aging seals may cause liquid working fluid leakage, aging pipes may cause increased circulation resistance or rupture, worn water pumps may cause insufficient circulation flow, and aging valves may cause control malfunction.
[0031] Therefore, the failure rate of the liquid cooling cycle system is not a constant value, but gradually increases with the increase of operating time. In this application, the third failure rate can be calculated based on the operating time of each key component in the liquid cooling cycle system, using the Weibull distribution or other distribution models suitable for aging failure, and then the instantaneous failure rates of each key component are added together to obtain the third failure rate.
[0032] The target failure rate is a quantitative assessment of the overall failure risk of a server system. The higher the value, the greater the likelihood that the server system will fail under its current operating conditions.
[0033] Step 104: In response to the target failure rate being greater than the preset failure rate threshold, determine the warning level based on the difference between the target failure rate and the preset failure rate threshold, and generate corresponding fault warning information and maintenance strategies based on the warning level.
[0034] In the embodiments of this application, the preset failure rate threshold is a pre-set safety boundary value for the overall failure rate of the server system, used to determine whether the current operational risk of the server system is within an acceptable range. When the target failure rate is less than or equal to the preset failure rate threshold, it indicates that the server system is currently operating well and no warning is required; when the target failure rate is greater than the preset failure rate threshold, it indicates that the overall failure risk of the server system has exceeded the safety boundary, and a warning needs to be issued to the maintenance personnel and corresponding maintenance recommendations need to be provided.
[0035] The warning level is a risk level based on the degree to which the target failure rate exceeds a preset failure rate threshold. In one embodiment of the present invention, the warning level can be divided into multiple levels according to the degree of exceedance, for example: yellow warning (first level warning), orange warning (second level warning) and red warning (third level warning), with the severity increasing sequentially.
[0036] Fault warning information is a notification message used to convey risk status to operation and maintenance personnel. It includes at least the current warning level, the target failure rate value, the specific extent to which the threshold is exceeded, and the main factors that cause the target failure rate to increase (e.g., whether the first failure rate of the liquid working fluid is too high, or the third failure rate of the liquid cooling circulation system is too high).
[0037] The maintenance strategy is a recommended maintenance plan generated based on the current warning level and failure cause analysis. It includes at least the suggested maintenance time window, the specific components or items to be maintained, and the priority order of maintenance operations. For example, when the primary failure rate is the main factor, the maintenance strategy may include suggestions such as starting the automatic maintenance system to filter and purify the liquid working fluid, replenishing with fresh liquid working fluid, or replacing all liquid working fluid; when the secondary failure rate is the main factor, the maintenance strategy may include checking and replacing seals, cleaning pipelines, or replacing the water pump.
[0038] Fault warning information and maintenance strategies can be output to the terminal equipment of maintenance personnel through the external interface (such as serial port, monitor or network port) of the central control system to guide maintenance personnel to take corresponding maintenance measures in a timely manner.
[0039] In some embodiments of the present invention, the generation process of fault warning information and maintenance strategies can be achieved using an artificial intelligence model. Specifically, a fault diagnosis and maintenance decision-making model can be pre-trained and deployed. This model takes data such as the target failure rate, the failure rate of each component, the warning level, and the system operation log as input, and outputs the optimal maintenance strategy. This model can be implemented using a case-based reasoning method or a rule-based reasoning method. For example, multiple maintenance rules can be pre-defined in the model, each rule corresponding to a combination of fault mode and warning level. When the input data matches a certain rule, the model outputs the maintenance strategy corresponding to that rule.
[0040] The server system management method of this application achieves dynamic prediction of the overall failure risk of the immersion liquid-cooled server system by real-time monitoring and quantitative evaluation of the physicochemical indicators of the liquid working fluid, combined with the failure rate calculation of the server and liquid cooling circulation system. This can improve the predictability of the server system operation risk, provide scientific decision-making basis for operation and maintenance personnel, thereby reducing the risk of unplanned downtime, extending the service life of equipment and reducing maintenance costs.
[0041] In one possible implementation of this application embodiment, the calculation of the first failure rate of the liquid working fluid can be achieved in the following ways, but is not limited to: determining multiple target correlation coefficients between the target physicochemical index and other physicochemical indexes, wherein the target physicochemical index is any one of the multiple physicochemical indexes; calculating the weighting coefficient corresponding to the target physicochemical index based on the target independent failure rate, the target preset temperature influence coefficient, and the multiple target correlation coefficients; and weighting and summing the independent failure rates corresponding to the multiple physicochemical indexes based on their respective weighting coefficients to obtain the first failure rate of the liquid working fluid.
[0042] In the embodiments of this application, the target physicochemical index refers to any one of the multiple physicochemical indexes. That is, in the cyclic calculation process of weighted processing of each physicochemical index, the physicochemical index currently selected as the processing object is defined as the target physicochemical index.
[0043] Other physicochemical indicators refer to the remaining physicochemical indicators among multiple physicochemical indicators other than the target physicochemical indicator. The target correlation coefficient is the correlation coefficient between the target physicochemical indicator and each of the other physicochemical indicators.
[0044] The target independent failure rate is the independent failure rate corresponding to the target physicochemical index. The target preset temperature influence coefficient is the temperature influence coefficient preset for the target physicochemical index. In this invention, the weighting coefficient is a proportional coefficient used to weight and sum the independent failure rates of each physicochemical index. It reflects the weight that should be assigned to the independent failure rate of each physicochemical index when comprehensively evaluating the overall failure rate of the liquid working fluid. This weighting coefficient depends on the correlation between the physicochemical index and other physicochemical indexes, as well as the sensitivity of the physicochemical index to temperature.
[0045] By iterating through each of the multiple physicochemical indicators and performing the above process for each, the weight coefficient corresponding to each physicochemical indicator can be obtained. Then, the independent failure rate of each physicochemical indicator is multiplied by the corresponding weight coefficient to obtain the weighted failure rate contribution value of each physicochemical indicator. Finally, the weighted failure rate contribution values of all physicochemical indicators are summed, and the sum is the first failure rate of the liquid working fluid.
[0046] This application uses a weighted summation method to give each physicochemical index an appropriate weight, so that the calculated first failure rate can accurately reflect the overall failure risk of the liquid working fluid under the current physicochemical index conditions.
[0047] This application overcomes the evaluation bias caused by directly adding the independent failure rates of various physicochemical indicators through correlation correction and temperature compensation, thereby improving the accuracy and reliability of liquid working fluid failure rate assessment.
[0048] In one possible implementation of this application embodiment, when calculating the weight coefficient corresponding to the target physicochemical index, it can be implemented in the following ways, but not limited to: summing the complements of each of the multiple target correlation coefficients to obtain a first coefficient; multiplying the target independent failure rate, the target preset temperature influence coefficient, and the first coefficient to obtain the target corrected failure rate corresponding to the target physicochemical index; and obtaining the weight coefficient corresponding to the target physicochemical index by dividing the target corrected failure rate by the sum of the corrected failure rates corresponding to each of the multiple physicochemical indexes, wherein the calculation method of the corrected failure rate corresponding to each of the multiple physicochemical indexes is the same as the calculation method of the target corrected failure rate.
[0049] In the embodiments of this application, the complement refers to the complementary value of a value relative to a preset base value. In one implementation, when the correlation coefficient ranges from [0,1], the complement of the correlation coefficient is defined as the difference between the correlation coefficient and the value 1, that is, the complement equals 1 minus the correlation coefficient. The correlation coefficient is used to characterize the strength and direction of the correlation between two physicochemical indicators. The closer the value is to 1, the stronger the positive correlation between the two physicochemical indicators, that is, the greater the probability that a change in one indicator will be accompanied by a change in the same direction as the other indicator; conversely, the closer the correlation coefficient is to 0, the weaker the linear correlation between the two physicochemical indicators, and the relatively independent changes in the two indicators. Therefore, the complement of the correlation coefficient (i.e., 1 minus the correlation coefficient) reflects the degree of uncorrelation or non-independence between the target physicochemical indicator and other corresponding physicochemical indicators.
[0050] The first coefficient is the cumulative value of the degree of uncorrelation between the target physicochemical indicator and all other physicochemical indicators. Its magnitude reflects the relative independence of the target physicochemical indicator among all monitored indicators. The specific operation of summing the complements of multiple target correlation coefficients is as follows: calculate the complement of each target correlation coefficient separately, and then add the calculated complements together. The sum is the first coefficient.
[0051] The target corrected failure rate is an intermediate calculated value obtained by correcting the independent failure rate of the target physicochemical index for both correlation and temperature. This value is used to determine the relative weight of the target physicochemical index in the overall evaluation of all physicochemical indexes.
[0052] Specifically, the target independent failure rate corresponding to the target physicochemical index, the target preset temperature influence coefficient set in advance for the target physicochemical index, and the first coefficient obtained by summing the complement of the correlation coefficient are multiplied together, and the resulting product is the target corrected failure rate.
[0053] Each of the multiple physicochemical indicators is calculated using the same method as the target physicochemical indicator. Specifically, for any given physicochemical indicator, its own value, its preset temperature influence coefficient, and the sum of the complements of its correlation coefficients with all other physicochemical indicators are multiplied to obtain the corresponding corrected failure rate. By iterating through all the indicators in the multiple physicochemical indicators, the corrected failure rate for each individual indicator can be obtained.
[0054] Finally, the target corrected failure rate of the target physicochemical index is used as the numerator, and the sum of the corrected failure rates of multiple physicochemical indices is used as the denominator. A division operation is then performed, and the quotient is the weighting coefficient corresponding to the target physicochemical index. The larger the corrected failure rate of a certain physicochemical index, the greater its impact on the overall failure rate of the liquid working fluid under the current state, and its weighting coefficient is correspondingly larger.
[0055] Specifically, the calculation of the weighting coefficients can be achieved using, but is not limited to, the following formula (1): Formula (1) in, The weighting coefficients corresponding to the target physicochemical indicators. Target independent failure rate, For any target, the correlation coefficient, For the number of physicochemical indicators, Set a temperature influence coefficient for the target. The independent failure rate corresponding to other physicochemical indicators, The correlation coefficients between other physicochemical indicators and different other physicochemical indicators. This is the preset temperature influence coefficient for other physicochemical indicators.
[0056] It should be noted that, This is the target corrected failure rate; This is the sum of the corrected failure rates corresponding to each of the multiple physicochemical indicators.
[0057] It should be further noted that the first failure rate of the liquid working fluid can be calculated using the following formula (2): Formula (2) in, This is the highest failure rate.
[0058] This application achieves adaptive determination of the weights of each physicochemical indicator through weight coefficient calculation. When the measured value of a certain physicochemical indicator deviates more from the standard value, has a stronger correlation with other physicochemical indicators, and is more sensitive to temperature changes, the corresponding corrected failure rate of that physicochemical indicator is greater, and thus it is assigned a larger weight coefficient. Conversely, when the measured value of a certain physicochemical indicator is close to the standard value, has a weaker correlation with other physicochemical indicators, and is not sensitive to temperature changes, its assigned weight coefficient is relatively small. This makes the calculation of the first failure rate of the liquid working fluid more reasonable and can accurately reflect the true impact of each physicochemical indicator under actual operating conditions.
[0059] In one possible implementation of this application embodiment, when obtaining the second failure rate of the server in the server system and the third failure rate of the liquid cooling circulation system, it can be implemented in the following ways, but not limited to: obtaining the server's runtime, calculating the second failure rate of the server based on the runtime and a first preset exponential distribution failure rate model; obtaining the component failure rates corresponding to multiple components in the liquid cooling circulation system, wherein the multiple components include at least one of pipes, valves, sealing rings and water pumps; and adding the component failure rates corresponding to the multiple components to obtain the third failure rate of the liquid cooling circulation system.
[0060] In the embodiments of this application, runtime refers to the cumulative working time of the server since it was put into operation, which can be measured in hours, days, or months. The first preset exponential distribution failure rate model is a pre-defined mathematical model used to describe the change in server failure rate over time. According to reliability engineering theory, the failure rate of electronic equipment usually follows an exponential distribution law during its service life, that is, the failure rate is approximately constant during the period of occasional failure of the equipment and does not change significantly with the increase of operating time. Based on this, the first preset exponential distribution failure rate model takes the server's runtime as the input parameter and the instantaneous server failure rate corresponding to the runtime as the output value.
[0061] Specifically, after obtaining the server's runtime, the first preset exponential distribution failure rate model is invoked, and the runtime is used as the model input for calculation to obtain the server's second failure rate at the current moment. It should be noted that in other embodiments, if the server has entered its wear-out and failure period, its failure rate may slowly increase over time. In this case, the parameters of the exponential distribution model can be modified based on the actual failure data, or other applicable distribution models can be used instead.
[0062] Next, regarding the acquisition of the third failure rate of the liquid cooling circulation system, the specific composition and quantity of multiple components can be determined according to the actual configuration of different server systems, and this invention does not impose specific limitations on this.
[0063] Component failure rate is the probability density of failure of each component at a given operating time. Its specific value is calculated based on the component type, accumulated operating time, and a pre-defined component failure model. In practice, the accumulated operating time is recorded for each component in the liquid cooling circulation system. For each component, a corresponding component failure rate calculation model can be pre-defined based on its failure mechanism. Given that the failure types of components such as pipes, valves, sealing rings, and pumps are mainly wear and aging failures, their failure rates gradually increase with operating time. This type of failure pattern can usually be described using the Weibull distribution.
[0064] The Weibull distribution is a probability distribution model suitable for describing the fatigue failure of mechanical parts and materials. Its failure rate function increases monotonically over time and can fit the time-dependent characteristics of aging and wear failures of parts.
[0065] Therefore, for each component, by substituting its cumulative runtime into the pre-stored Weibull distribution failure rate model, the current instantaneous failure rate of that component can be calculated, which is the component's failure rate. The shape and scale parameters of the Weibull distribution model corresponding to each component can be predetermined based on the component's material properties, operating environment stress level, and historical failure statistics.
[0066] Since the failure of any component in a liquid cooling cycle system will lead to the loss of the overall function or performance degradation of the liquid cooling cycle system, the overall failure rate of the liquid cooling cycle system is equal to the sum of the failure rates of all its components. After obtaining the individual component failure rates, the component failure rates are summed, and the resulting sum is the third failure rate of the liquid cooling cycle system.
[0067] This application describes the failure rate variation patterns of each component of the server equipment and liquid cooling circulation system by using exponential distribution and Weibull distribution respectively, fully considering the essential differences in failure mechanisms of different equipment types, thereby making the calculation of the second and third failure rates more consistent with actual operating conditions.
[0068] In one possible implementation of this application embodiment, when calculating the independent failure rate corresponding to each of the multiple physicochemical indicators, the following method can be used, but is not limited to: calculating the independent failure rate corresponding to each of the multiple physicochemical indicators based on the indicator values and the second preset exponential distribution failure rate model.
[0069] In the embodiments of this application, the second preset exponential distribution failure rate model is a pre-defined mathematical model used to describe the mapping relationship between the index value of a single physicochemical indicator and the failure rate of the liquid working fluid caused independently by that physicochemical indicator. Unlike the first preset exponential distribution failure rate model used in the calculation of the second failure rate, the independent variable of the second preset exponential distribution failure rate model is the index value of the physicochemical indicator rather than the running time, and its dependent variable is the independent failure rate corresponding to that physicochemical indicator.
[0070] In one possible implementation of this application embodiment, when determining the warning level based on the difference between the target failure rate and the preset failure rate threshold, the following methods can be used, but are not limited to: calculating the target difference between the target failure rate and the preset failure rate threshold, and calculating the target ratio between the target difference and the preset failure rate threshold; determining the warning level as a first-level warning in response to the target ratio being within a first preset value interval; determining the warning level as a second-level warning in response to the target ratio being within a second preset value interval; and determining the warning level as a third-level warning in response to the target ratio being within a third preset value interval; wherein the first, second, and third preset value intervals do not overlap, the minimum value in the third preset value interval is greater than the maximum value in the second preset value interval, and the minimum value in the second preset value interval is greater than the maximum value in the first preset value interval.
[0071] In the embodiments of this application, the target difference refers to the difference obtained by subtracting the preset failure rate threshold from the target failure rate. This difference reflects the specific extent to which the overall failure rate of the current server system exceeds the safety boundary. The target ratio refers to the quotient obtained by dividing the target difference by the preset failure rate threshold. This ratio represents the relative degree to which the target failure rate exceeds the safety boundary in the form of a relative percentage.
[0072] The first preset numerical range corresponds to the lowest risk warning level, with a minimum value greater than zero and a maximum value less than the minimum value of the second preset numerical range. When the calculated target ratio falls within the first preset numerical range, it indicates that the target failure rate of the server system exceeds the preset failure rate threshold by a relatively small margin, and the operational risk is in a state of mild over-limit. At this time, the warning level is determined to be Level 1. Level 1 warning indicates that the server system currently has a mild risk. Maintenance personnel should pay attention to the system's operating status and may arrange inspections or maintenance work as appropriate, but there is no need to take immediate emergency measures.
[0073] The second preset numerical range corresponds to a moderate risk warning level. Its minimum value is greater than the maximum value of the first preset numerical range, and its maximum value is less than the minimum value of the third preset numerical range. When the calculated target ratio falls within the second preset numerical range, it indicates that the relative extent to which the target failure rate of the server system exceeds the preset failure rate threshold is at a moderate level, and the operational risk has clearly exceeded the limit. At this time, the warning level is determined to be the second-level warning. The second-level warning indicates that the server system has a moderate risk, and maintenance personnel should arrange targeted maintenance work as soon as possible to reduce the overall failure rate of the server system and prevent the risk from further deteriorating.
[0074] The third preset numerical range corresponds to the highest risk warning level, with its minimum value exceeding the maximum value of the second preset numerical range. When the calculated target ratio falls within the third preset numerical range, it indicates that the target failure rate of the server system exceeds the preset failure rate threshold by a relatively large margin, and the operational risk is in a severely excessive state. At this point, the warning level is determined to be Level 3. A Level 3 warning indicates that the server system faces serious risks, and a functional failure may occur at any time. Maintenance personnel should immediately take measures such as shutdown for repair or other emergency measures to prevent major equipment damage or data loss.
[0075] It should be noted that any value of the target ratio falls within at most one preset numerical range, thus ensuring that each warning judgment corresponds to a unique and definite warning level and avoiding ambiguity. The specific endpoint values of the three preset numerical ranges can be preset according to factors such as the server system's reliability requirements, operational response capabilities, and historical operating data, and can be dynamically adjusted based on actual operating experience.
[0076] In one possible implementation, the first-level warning is a yellow warning, the second-level warning is an orange warning, and the third-level warning is a red warning, with the severity of the three warning levels increasing sequentially. Accordingly, after determining the warning level, the generated fault warning information must at least include identification information for the current warning level (e.g., warning color, warning level number, or warning text description) so that maintenance personnel can quickly and intuitively understand the severity of the current risk situation. Simultaneously, different warning levels can correspond to different information push methods and response processes. For example, a first-level warning can be notified to maintenance personnel via status indicator lights on the system interface or through periodic summary reports; a second-level warning can be sent to maintenance personnel via SMS or email; and a third-level warning can convey risk information to maintenance personnel through emergency methods such as audible and visual alarms, telephone calls, or automatic triggering of backup system switching.
[0077] Furthermore, after determining the warning level, the maintenance strategy can be further adjusted based on the level of the warning. For example, when the warning level is Level 1, the recommended maintenance execution time window in the maintenance strategy is relatively flexible and can be arranged to be handled uniformly within the regular maintenance cycle; when the warning level is Level 2, the recommended maintenance execution time window in the maintenance strategy is appropriately shortened, and it is recommended to complete the maintenance work within the predetermined time; when the warning level is Level 3, the recommended maintenance execution time window in the maintenance strategy is the shortest, and it is recommended to execute the maintenance work immediately.
[0078] This application uses the target ratio of the calculated target difference to a preset failure rate threshold as the basis for determining the warning level. It also sets three non-overlapping preset numerical ranges corresponding to three different severity warning levels, thus achieving graded quantitative assessment and graded warning of server system failure risks. Maintenance personnel can quickly determine the severity of the risk based on the warning level and take corresponding countermeasures, improving the precision of server system operation and maintenance management and emergency response efficiency.
[0079] In one possible implementation of this application embodiment, the fault risk level of the first-level warning indication is lower than the fault risk level of the second-level warning indication; The fault risk level of a Level 2 warning indication is lower than that of a Level 3 warning indication.
[0080] Among them, the fault risk level is an indicator used to characterize the severity of the current operational risk of the server system.
[0081] In one possible implementation of this application embodiment, in summary of the above embodiments, to facilitate understanding of the server system fault warning process, this application embodiment also provides a schematic diagram of the server system warning process, such as... Figure 2 As shown, the central control system of the server system calculates and outputs the dynamic failure rate of the immersion liquid-cooled server in real time, i.e., the expected failure rate. When the failure rate of a single component or a single indicator exceeds its own failure rate threshold, a failure risk warning for that single component or single indicator is issued; when the expected failure rate of the entire system is higher than the failure rate threshold of the entire system, a system failure risk warning is issued. Based on the percentage by which the expected failure rate exceeds the failure rate threshold, the failure risk warning level is defined as yellow, orange, and red, with severity ranging from low to high. When a single component or multiple components issue warnings of different degrees, the AI model in the central control system can formulate the optimal maintenance plan for the immersion liquid-cooled server based on factors such as the severity of the warning, the availability of spare parts, and the maintenance cycle.
[0082] In one possible implementation of this application embodiment, after acquiring the index values corresponding to multiple physicochemical indicators of the liquid working fluid in the server system in real time, the following methods can also be used, but are not limited to: controlling the automatic maintenance system to perform maintenance processing on the liquid working fluid according to the index values; the maintenance processing includes at least one of filtering solid impurities in the liquid working fluid, adsorbing free ions in the liquid working fluid, and absorbing water molecules in the liquid working fluid.
[0083] In the embodiments of this application, the automatic maintenance system refers to the equipment assembly installed in the server system for automatically purifying liquid working fluids. It works in conjunction with the detection system and the central control system to form the automatic maintenance architecture of the server system.
[0084] Specifically, filtering solid impurities from a liquid working medium refers to the operation of separating solid particles suspended or dispersed in the liquid working medium from the liquid working medium using physical filtration methods. Adsorbing free ions from a liquid working medium refers to the operation of removing free charged ions from the liquid working medium using ion exchange or physical adsorption methods. Absorbing water molecules from a liquid working medium refers to the operation of removing dissolved or mixed water from the liquid working medium using adsorption or separation methods.
[0085] In practice, based on the specific types and exceedances of various physicochemical indicators acquired in real time, the corresponding devices in the automatic maintenance system are selectively activated. For example, when the kinematic viscosity of the liquid working fluid is detected to increase beyond the standard value, it usually indicates the presence of a large amount of solid particulate impurities. In this case, a command is sent to the automatic maintenance system to activate the solid particle filtration device to filter the liquid working fluid. When the volume resistivity or breakdown voltage of the liquid working fluid is detected to decrease, it usually indicates the presence of a large amount of free ions or water. In this case, based on preset judgment logic or combined with the detection results of other indicators, the main pollutant type is determined, and the ion filtration device and / or water molecule adsorption device are activated accordingly for purification.
[0086] This application realizes automatic online maintenance of liquid working fluid during the operation of the server system. Specifically, when the physical and chemical properties of the liquid working fluid are detected to deteriorate, the corresponding purification device is automatically activated to intervene, so that the physical and chemical properties of the liquid working fluid can be repaired in time after deviating from the standard value. This slows down the overall deterioration process of the liquid working fluid, extends its service life, reduces the replacement frequency of the liquid working fluid, and lowers the system's operation and maintenance costs.
[0087] In one possible implementation of this application embodiment, when controlling the automatic maintenance system to perform maintenance on the liquid working fluid according to the index value, it can be implemented in the following ways, but not limited to: comparing the index value corresponding to each of the multiple physicochemical indicators with the preset index threshold corresponding to each of the multiple physicochemical indicators to obtain the comparison result; in response to the comparison result that at least one physicochemical indicator has an index value greater than the corresponding preset index threshold and the number of automatic maintenance is less than the preset number threshold, accumulating the number of automatic maintenance, and controlling the automatic maintenance system to perform maintenance on the liquid working fluid.
[0088] In the embodiments of this application, the preset index threshold is a pre-set judgment boundary value for each physicochemical index, used to measure whether the current index value of the physicochemical index is within an acceptable normal working range. It should be noted that for different types of physicochemical indices, the preset index threshold can be set as the upper limit (i.e., the maximum allowable value) and / or the lower limit (i.e., the minimum allowable value) of the standard value range. The specific setting method depends on the failure mechanism of the physicochemical index in the liquid working fluid. For example, for the kinematic viscosity index, when the kinematic viscosity is too high, it will significantly affect the fluidity and heat dissipation efficiency of the liquid working fluid; therefore, an upper limit threshold for kinematic viscosity needs to be set as the preset index threshold. For the volume resistivity index, when the volume resistivity is too low, it will reduce the insulation performance of the liquid working fluid; therefore, a lower limit threshold for volume resistivity needs to be set as the preset index threshold. The preset index threshold corresponding to each physicochemical index is pre-determined based on the type of liquid working fluid, the design requirements of the server system, and relevant safety specifications.
[0089] The comparison result refers to the determination of the magnitude relationship obtained by comparing the current value of each physicochemical indicator with the preset threshold value corresponding to that physicochemical indicator. The comparison result includes at least information on whether the value of each physicochemical indicator exceeds its corresponding preset threshold value.
[0090] In this invention, the automatic maintenance count is used to record the cumulative number of times the automatic maintenance system has performed maintenance on the liquid working fluid since the last maintenance count was reset to zero. Each time the automatic maintenance system completes a full maintenance operation, the automatic maintenance count is updated accordingly. The preset count threshold is a pre-set maximum allowable value for the number of automatic maintenance operations, used to limit the maximum number of automatic maintenance operations allowed within a single maintenance cycle. The specific value of the preset count threshold can be determined comprehensively based on factors such as the processing capacity of each purification device in the automatic maintenance system, the total capacity of the liquid working fluid, and the purifiable and reversible characteristics of the liquid working fluid under different levels of contamination. For example, 3 times means that the maximum number of automatic maintenance operations allowed is 3. However, this value is only illustrative and can be set to other values according to actual needs in other embodiments; this invention does not specifically limit this.
[0091] Specifically, after obtaining the aforementioned comparison results, it is determined whether at least one physicochemical indicator in the comparison results has a value greater than its corresponding preset threshold. If the values of all physicochemical indicators do not exceed their respective preset thresholds, it indicates that all physicochemical performance indicators of the liquid working fluid are within the qualified range, and no maintenance is required; the current state should be maintained and monitoring should continue. If the value of at least one physicochemical indicator in the comparison results exceeds its corresponding preset threshold, it indicates that at least one physicochemical indicator in the liquid working fluid has exceeded the standard, and purification treatment of the liquid working fluid is required.
[0092] If the number of automatic maintenance attempts is less than the preset threshold, it indicates that the automatic maintenance system still has capacity to perform maintenance. In this case, an accumulation operation is performed: the current count of automatic maintenance attempts is incremented by 1, and the updated count is stored in the storage unit to record this maintenance trigger event. Subsequently, based on the specific type of the physicochemical indicator exceeding the preset threshold, the type of maintenance process to be initiated is determined, and corresponding control commands are issued to the automatic maintenance system. For example, if the physicochemical indicator exceeding the threshold is kinematic viscosity, the central control system determines that solid impurity filtration of the liquid working fluid is required and issues a start command to the solid particle filtration device in the automatic maintenance system; if the physicochemical indicator exceeding the threshold is volume resistivity or breakdown voltage, the central control system determines that ion adsorption and / or water molecule absorption of the liquid working fluid is required and issues start commands to the ion filtration device and / or water molecule adsorption device in the automatic maintenance system.
[0093] If the number of automatic maintenance operations has reached or exceeded the preset threshold, it indicates that the automatic maintenance system has performed the preset allowed number of maintenance operations. Continuing to perform maintenance operations can no longer effectively improve the physicochemical properties of the liquid working fluid. At this time, the cumulative operation and control maintenance operations will no longer be performed. Instead, a corresponding alarm will be triggered (e.g., automatic maintenance purification failure alarm) to notify the operation and maintenance personnel that the liquid working fluid can no longer be restored to a qualified state through automatic maintenance and manual intervention measures are required, such as replacing the liquid working fluid or performing deep maintenance.
[0094] This application ensures that the quality of the liquid working fluid remains within an acceptable range by promptly triggering maintenance procedures when any physicochemical indicator exceeds the standard, thus avoiding the risk of system failure due to continuous deterioration of the liquid working fluid's performance. By setting a preset threshold to limit the number of automatic maintenance operations, it avoids repeatedly performing ineffective maintenance operations when the liquid working fluid is severely contaminated and the automatic maintenance device can no longer effectively purify it, saving system resources and promptly detecting irreversible failure states of the liquid working fluid and triggering alarms, thereby ensuring the reliability and security of the server system.
[0095] In one possible implementation of this application embodiment, after controlling the automatic maintenance system to perform maintenance on the liquid working fluid, the following methods may also be used, but are not limited to: re-acquiring the post-maintenance index values corresponding to each of the multiple physicochemical indicators; in response to the fact that the post-maintenance index values corresponding to each of the multiple physicochemical indicators are not greater than the corresponding preset index thresholds, the automatic maintenance count is reset to zero; in response to the fact that the post-maintenance index value of at least one physicochemical indicator is greater than the corresponding preset index threshold, the automatic maintenance count is accumulated, and the automatic maintenance system is controlled again to perform maintenance on the liquid working fluid; in response to the automatic maintenance count being greater than the preset count threshold, purification failure information is output.
[0096] In the embodiments of this application, the post-maintenance index value refers to the real-time index value obtained by the detection system after the automatic maintenance system has completed the maintenance operation and the detection system has re-detected the various physicochemical indicators of the liquid working fluid.
[0097] After obtaining the post-maintenance values of each physical and chemical indicator, the post-maintenance value of each physical and chemical indicator is compared with the preset threshold value corresponding to that physical and chemical indicator, and the following different processes are performed according to the comparison results.
[0098] First, if it is confirmed that the post-maintenance values of each physicochemical indicator do not exceed their corresponding preset thresholds, it indicates that the maintenance process has successfully restored all out-of-range physicochemical indicators of the liquid working fluid to within the acceptable range, and the various physicochemical performance indicators of the liquid working fluid meet the normal operation requirements of the server system. At this point, in response to this determination, the automatic maintenance count is reset to zero. In this invention, resetting means resetting the current count of automatic maintenance to zero.
[0099] Second, if it is confirmed that the post-maintenance values of one or more physicochemical indicators still exceed their corresponding preset thresholds after maintenance, it indicates that although the maintenance has been performed, some physicochemical indicators in the liquid working medium have not yet returned to the acceptable range, and further maintenance is required. In response to this determination, the automatic maintenance count is incremented, and the automatic maintenance system is re-controlled to perform maintenance on the liquid working medium. Specifically, the count of the currently recorded automatic maintenance count is incremented by 1 to record this maintenance attempt, and then a control command is sent to the automatic maintenance system again to restart the corresponding maintenance and purification device to perform a new round of maintenance on the liquid working medium. After the new round of maintenance is completed, the post-maintenance values of each physicochemical indicator are re-acquired and judged, i.e., the above verification and iteration process is repeated until the post-maintenance values of all physicochemical indicators are no greater than their corresponding preset thresholds, or the number of automatic maintenance operations reaches the preset threshold.
[0100] Third, during the aforementioned iterative cycle, after each accumulation of automatic maintenance counts, the accumulated automatic maintenance count is compared with a preset threshold. When the automatic maintenance count exceeds the preset threshold, it indicates that the automatic maintenance system has performed maintenance processing exceeding the preset allowed number of times, but at least one physicochemical indicator in the liquid working fluid still fails to return to the acceptable range. At this point, the maintenance system is unable to restore the liquid working fluid to a state that meets operational requirements through its own purification device, and continuing to perform maintenance processing is meaningless. In response to this determination result, the operation of outputting purification failure information is performed. In this invention, purification failure information is a notification message used to inform maintenance personnel that the liquid working fluid cannot be restored to a qualified state through automatic maintenance. Its specific content includes at least: the specific physicochemical indicator type that cannot be purified and its current post-maintenance indicator value, the type and number of maintenance processes performed, and suggested follow-up measures for maintenance personnel (e.g., replacing the liquid working fluid, checking whether the working status of each device in the automatic maintenance system is normal, or performing manual deep maintenance on the server system, etc.).
[0101] Cleanup failure information can be output to the terminal equipment of maintenance personnel through the external interface of the central control system (such as serial port, monitor or network port) in the form of text display, status indicator light, email or SMS, so that maintenance personnel can be informed in time and take corresponding manual intervention measures.
[0102] This application accurately assesses the effectiveness of maintenance by re-testing various physicochemical indicators of the liquid working fluid after maintenance and comparing them with preset threshold values, thus avoiding the risk of continuing operation with the liquid working fluid incompletely restored to a qualified state. Through iterative maintenance, the automated maintenance system is allowed to attempt purification multiple times when a single maintenance attempt is insufficient, extending the service life of the liquid working fluid and reducing the frequency of manual intervention. By promptly outputting purification failure information when the number of automated maintenance attempts exceeds the limit, it ensures that irreversible failures of the liquid working fluid can be detected and notified to maintenance personnel in a timely manner, improving the security and operational efficiency of the server system.
[0103] In one possible implementation of this application embodiment, when controlling the automatic maintenance system to perform maintenance on the liquid working fluid, it can be implemented in the following ways, but not limited to: starting the maintenance device corresponding to the physical and chemical index to be purified in the automatic maintenance system to perform maintenance, wherein the physical and chemical index to be purified is a physical and chemical index whose index value is greater than the corresponding preset index threshold.
[0104] In the embodiments of this application, the physicochemical index to be purified refers to the physicochemical index among multiple physicochemical indexes whose currently acquired real-time index value is greater than its corresponding preset index threshold. The physicochemical index to be purified is the physicochemical index that has exceeded the qualified range and needs to be restored through maintenance. Since the deterioration of different physicochemical indexes is usually caused by different types of impurities, it is necessary to selectively activate the corresponding maintenance device according to the specific type of the physicochemical index to be purified in order to achieve precise maintenance.
[0105] In this invention, a maintenance device is a physical equipment unit in an automated maintenance system used to perform specific types of impurity removal operations. The automated maintenance system includes various types of maintenance devices, each corresponding to different physicochemical purification requirements. Specifically, the maintenance devices include at least a solid particle filter, an ion filter, and a water molecule adsorption device. A pre-established correspondence between each maintenance device and the physicochemical indicator is established. During implementation, when a certain physicochemical indicator is identified as requiring purification, the type of maintenance device corresponding to that indicator is determined based on the correspondence, and a start command is issued to that type of maintenance device.
[0106] When multiple physicochemical indicators in the comparison results are found to have values greater than their respective preset thresholds, indicating multiple physicochemical indicators to be purified, the corresponding maintenance device for each indicator is determined, and a start command is issued to all determined maintenance devices. Upon receiving the start command, each maintenance device starts operating in parallel or sequentially, performing its corresponding maintenance processing operation.
[0107] This application only activates the maintenance devices corresponding to the actual physicochemical indicators that exceed the standards, without activating other unrelated maintenance devices. This avoids unnecessary idling of maintenance devices, reduces energy consumption and wear of the automatic maintenance system, extends the service life of the maintenance devices, and the targeted maintenance treatment can more efficiently remove specific types of impurities that cause the physicochemical indicators to exceed the standards, avoiding the low treatment efficiency caused by blindly carrying out comprehensive purification.
[0108] In one possible implementation of this application embodiment, after generating corresponding fault warning information and maintenance strategy according to the warning level, the following method can also be used, but is not limited to: transmitting the target failure rate, fault warning information and maintenance strategy to external devices through an external interface.
[0109] In the embodiments of this application, the external interface is a physical or logical port used for data communication with external devices. It is used to output data, information, and instructions generated internally by the system to external devices for viewing, analysis, and processing by maintenance personnel or other systems. The specific form of the external interface can be selected according to the actual configuration and maintenance management needs of the server system.
[0110] External devices refer to terminal devices that are independent of the system and capable of receiving and processing information output by the system.
[0111] In one possible implementation of this application embodiment, the plurality of physicochemical indicators include at least one of kinematic viscosity, electrical properties, pH, and thermal performance indicators; Electrical properties include at least one of dielectric constant, volume resistivity, and breakdown voltage; Thermal performance indicators include at least one of specific heat and thermal conductivity.
[0112] In one possible implementation of this application, the correlation coefficient includes at least the Pearson correlation coefficient.
[0113] In one possible implementation of this application embodiment, to facilitate understanding of the automatic detection and maintenance process of the liquid working fluid, this application embodiment also provides a schematic diagram of the automatic maintenance process, such as... Figure 3 As shown, the detection system monitors the various physicochemical indicators of the liquid working fluid in real time and feeds the results back to the central control system. The central control system stores threshold values for each physicochemical indicator of the liquid working fluid and analyzes the detection results from the detection system, comparing the detected values with the thresholds. If the detection result exceeds the threshold range, the central control system issues an alarm for excessive physicochemical indicators of the liquid working fluid based on the type of indicator exceeding the threshold range and automatically increments the maintenance count by 1. When the automatic maintenance count is ≤3, the central control system issues a command to activate the corresponding maintenance and purification device. After the maintenance action is completed, the central control system collects the detection values of the physicochemical indicators of the liquid working fluid from the detection system again. If the physicochemical indicators are qualified, the automatic maintenance count is reset to zero, and the process ends; if the automatic maintenance count is >3, it is determined that the physicochemical indicators of the liquid working fluid cannot be purified by the automatic maintenance device, and an automatic maintenance and purification failure alarm is issued, ending the process.
[0114] Embodiments of this application also provide a server system. Figure 4 A schematic diagram of the structure of a server system provided in this application is shown below. Figure 4 As shown, it includes: The detection system is used to detect the values of multiple physicochemical indicators of the liquid working fluid in the server system in real time. An automated maintenance system is used to perform maintenance on liquid working fluids. The central control system is connected to both the detection system and the automatic maintenance system. The central control system is used to perform tasks such as... Figure 1 Management methods for server systems in China.
[0115] The detection system is responsible for detecting and collecting various physicochemical properties of the liquid working fluid and feeding the results back to the central control system. The detection system consists of the following components: an integrated circuit (IC), a kinematic viscosity testing device, an electrical performance testing device, a pH value testing device, and a thermal performance testing device.
[0116] The detection system IC is responsible for monitoring the status of each detection subsystem, collecting the detection information and results from each detection subsystem, and summarizing and feeding them back to the central control system.
[0117] Kinematic viscosity testing device: for detecting the kinematic viscosity of liquid working fluids.
[0118] Electrical performance testing equipment: tests electrical parameters such as dielectric constant, volume resistivity, and breakdown voltage of liquid working fluids.
[0119] pH value detection device: detects the pH value of liquid working fluid.
[0120] Thermal performance testing device: used to test thermal performance indicators such as specific heat and thermal conductivity of liquid working fluids.
[0121] The automatic maintenance system is responsible for filtering and removing impurities from the liquid working fluid to maintain, to some extent, the physicochemical properties of the liquid working fluid meeting the requirements for the overall system operation. The automatic maintenance system consists of the following components: an automatic maintenance system IC, a solid particle filter, an ion filter, and a water molecule adsorption device.
[0122] Automatic maintenance system IC: It is responsible for monitoring the status of each maintenance subsystem, receiving instructions from the central control system and controlling each maintenance subsystem to complete the liquid working fluid cleaning work, and summarizing and feeding back the status information and maintenance results of the automatic maintenance system to the central control system.
[0123] Solid particle filtration device: responsible for filtering and adsorbing various solid impurities in liquid working fluids.
[0124] Ion filtration device: responsible for adsorbing free ions in liquid working fluid.
[0125] Water molecule adsorption device: responsible for absorbing water molecules in the working fluid.
[0126] The central control system, as the control center of the overall system, has the following main functions: Collect, analyze, and process information fed back from detection systems and automated maintenance systems; Send instructions to the detection system and automatic maintenance system, instructing them to complete the detection of the physical and chemical properties of the liquid working fluid and the automatic maintenance actions (filtering and removing impurities from the liquid working fluid). Calculate the expected failure rate of the overall system, compare the expected failure rate with the failure rate threshold, and output the expected failure rate, failure risk warning (comparison result of expected failure rate and threshold) and maintenance plan through external interfaces (serial port, monitor, network port).
[0127] For a description of the features in the embodiment corresponding to the server system, please refer to the relevant description of the embodiment corresponding to the management method of the server system, which will not be repeated here.
[0128] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method.
[0129] Embodiments of this application also provide an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to perform the steps in any of the above-described server system management method embodiments.
[0130] Embodiments of this application also provide a computer-readable storage medium storing a computer program, wherein the computer program is configured to execute the steps in any of the above-described server system management method embodiments when it runs.
[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 computer program product, which includes a computer program that, when executed by a processor, implements the steps in any of the above-described server system management method embodiments.
[0133] Embodiments of this application also provide another computer program product, including a non-volatile computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps in any of the above-described server system management method embodiments.
[0134] Any of the components, modules, units, parts, methods, and operations described herein can be implemented using software, firmware, hardware (e.g., fixed logic circuitry), manual processing, or any combination thereof. Alternatively or additionally, any functionality described herein can be executed at least in part by one or more hardware logic components, such as, but not limited to, a central processing unit (CPU), a field-programmable gate array (FPGA), an application-specific integrated circuit (ASIC), an application-specific standard product (ASSP), a system-on-a-chip (SoC), a complex programmable logic device (CPLD), a microprocessor (MCU), etc. The terms "system," "computing device," or "apparatus" as used herein encompass various means, devices, and machines for processing data, including, for example, one or more programmable processors, computers, SoCs, or combinations thereof. The apparatus may also include code that creates an execution environment for the computer program in question, such as code constituting processor firmware, a protocol stack, a database management system, an operating system, a cross-platform runtime environment, a virtual machine, or one or more combinations thereof. The aforementioned computer program (also known as a program, software, software application, app, script, or code) can be written in any form of programming language, including compiled or interpreted languages, declarative or procedural languages, and can be deployed in any form, including as a standalone program or as a module, component, subroutine, object, or other unit suitable for a computing environment.
[0135] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction 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, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0136] The foregoing has provided a detailed description of a server system management method and the server system itself. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are merely for the purpose of helping to understand the method and its core ideas. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the protection scope of the claims of this application.
Claims
1. A management method of a server system, characterized by, include: The system acquires the values of multiple physicochemical indicators of the liquid working fluid in the server system in real time, and calculates the independent failure rate of each of the multiple physicochemical indicators based on the indicator values. The independent failure rate is the failure rate of the liquid working fluid caused by the corresponding physicochemical indicator. Determine the correlation coefficients among the multiple physicochemical indicators, and calculate the independent failure rate by weighting the independent failure rate based on the correlation coefficients and the preset temperature influence coefficients corresponding to each of the multiple physicochemical indicators, to obtain the first failure rate of the liquid working fluid. The second failure rate of the server and the third failure rate of the liquid cooling circulation system in the server system are obtained. The second failure rate, the third failure rate and the first failure rate are added together to obtain the target failure rate of the server system. In response to the target failure rate being greater than a preset failure rate threshold, a warning level is determined based on the difference between the target failure rate and the preset failure rate threshold, and corresponding fault warning information and maintenance strategies are generated based on the warning level. The weighted calculation of the independent failure rate based on the correlation coefficient and the preset temperature influence coefficient corresponding to each of the multiple physicochemical indicators includes: Determine multiple target correlation coefficients between the target physicochemical index and other physicochemical indicators, wherein the target physicochemical index is any one of the multiple physicochemical indicators; Based on the target independent failure rate, the target preset temperature influence coefficient, and the multiple target correlation coefficients corresponding to the target physicochemical indicators, calculate the weighting coefficients corresponding to the target physicochemical indicators. Based on the weighting coefficients corresponding to the multiple physicochemical indicators, the independent failure rates corresponding to the multiple physicochemical indicators are weighted and summed to obtain the first failure rate of the liquid working fluid. The calculation of the weighting coefficients corresponding to the target physicochemical indicators includes: The first coefficient is obtained by summing the complements of the multiple target correlation coefficients. The target independent failure rate, the target preset temperature influence coefficient, and the first coefficient are multiplied to obtain the target corrected failure rate corresponding to the target physicochemical index. The weighting coefficients corresponding to the target physicochemical indicators are obtained by dividing the target corrected failure rate by the sum of the corrected failure rates corresponding to each of the plurality of physicochemical indicators. The calculation method of the corrected failure rate corresponding to each of the plurality of physicochemical indicators is the same as the calculation method of the target corrected failure rate.
2. The management method of a server system according to claim 1, characterized by, The process of obtaining the second failure rate of the server and the third failure rate of the liquid cooling circulation system in the server system includes: Obtain the runtime of the server, and calculate the second failure rate of the server based on the runtime and the first preset exponential distribution failure rate model; Obtain the component failure rate of each of the multiple components in the liquid cooling circulation system, wherein the multiple components include at least one of pipes, valves, sealing rings and water pumps; The failure rates of each of the multiple components are added together to obtain the third failure rate of the liquid cooling cycle system.
3. The management method of a server system according to claim 1, characterized by, The calculation of the independent failure rate corresponding to each of the multiple physicochemical indicators based on the indicator values includes: Based on the index values and the second preset index distribution failure rate model, calculate the independent failure rate corresponding to each of the multiple physicochemical indices.
4. The management method of a server system according to claim 1, characterized by, The step of determining the warning level based on the difference between the target failure rate and the preset failure rate threshold includes: Calculate the target difference between the target failure rate and the preset failure rate threshold, and calculate the target ratio between the target difference and the preset failure rate threshold; In response to the target ratio being within a first preset numerical range, the warning level is determined to be a first-level warning. In response to the target ratio being within a second preset numerical range, the warning level is determined to be a second-level warning. In response to the target ratio being within a third preset numerical range, the warning level is determined to be a third-level warning. The first preset value interval, the second preset value interval, and the third preset value interval do not overlap. The minimum value in the third preset value interval is greater than the maximum value in the second preset value interval, and the minimum value in the second preset value interval is greater than the maximum value in the first preset value interval.
5. The management method of a server system according to claim 4, wherein The fault risk level of the first-level warning indication is lower than that of the second-level warning indication. The fault risk level of the second-level warning indication is lower than that of the third-level warning indication.
6. The management method of a server system according to claim 1, wherein After acquiring the values of multiple physicochemical indicators of the liquid working fluid in the server system in real time, the method further includes: The automatic maintenance system is controlled to perform maintenance procedures on the liquid working fluid based on the aforementioned index values. The maintenance process includes at least one of filtering solid impurities from the liquid working medium, adsorbing free ions from the liquid working medium, and absorbing water molecules from the liquid working medium.
7. The management method of a server system according to claim 6, wherein The step of controlling the automatic maintenance system to perform maintenance on the liquid working fluid according to the index value includes: The index values corresponding to each of the multiple physicochemical indicators are compared with the preset index thresholds corresponding to each of the multiple physicochemical indicators to obtain the comparison results; In response to the fact that at least one physicochemical index value in the comparison result is greater than the corresponding preset index threshold, and the number of automatic maintenance is less than the preset number threshold, the number of automatic maintenance is accumulated, and the automatic maintenance system is controlled to perform maintenance processing on the liquid working fluid.
8. The server system management method according to claim 7, characterized in that, After the automatic maintenance system performs maintenance on the liquid working fluid, the method further includes: Reacquire the post-maintenance index values corresponding to each of the multiple physicochemical indicators; If the post-maintenance values of the multiple physical and chemical indicators are all no greater than the corresponding preset indicator thresholds, then the number of automatic maintenance operations will be reset to zero. In response to the fact that at least one physical or chemical indicator value is greater than the corresponding preset indicator threshold after maintenance, the number of automatic maintenance times is accumulated, and the automatic maintenance system is re-controlled to perform maintenance on the liquid working fluid. If the number of automatic maintenance operations exceeds the preset threshold, a purification failure message is output.
9. The server system management method according to claim 7, characterized in that, The automatic maintenance system performs maintenance procedures on the liquid working fluid, including: The maintenance device corresponding to the physical and chemical index to be purified in the automatic maintenance system is activated to perform maintenance processing, wherein the physical and chemical index to be purified is a physical and chemical index whose index value is greater than the corresponding preset index threshold.
10. The server system management method according to claim 1, characterized in that, After generating corresponding fault warning information and maintenance strategies based on the warning level, the method further includes: The target failure rate, the fault warning information, and the maintenance strategy are transmitted to external devices through an external interface.
11. The server system management method according to claim 1, characterized in that, The plurality of physicochemical properties include at least one of kinematic viscosity, electrical properties, pH, and thermal performance properties; The electrical properties include at least one of dielectric constant, volume resistivity, and breakdown voltage. The thermal performance indicators include at least one of specific heat and thermal conductivity.
12. The server system management method according to claim 1, characterized in that, The correlation coefficients include at least the Pearson correlation coefficient.
13. A server system, characterized in that, include: The detection system is used to detect the values of multiple physicochemical indicators of the liquid working fluid in the server system in real time. An automatic maintenance system is used to perform maintenance on the liquid working fluid; A central control system is connected to the detection system and the automatic maintenance system respectively, and the central control system is used to execute the management method of the server system as described in any one of claims 1 to 12.
Citation Information
Patent Citations
Intelligent comprehensive management method and system for liquid cooling system and server
CN117492544A
Detection service method and system based on artificial intelligence and detection data
CN121281215A
