Data center operating temperature monitoring method, cabinet door detection method, and monitoring system

Abstract

The present application relates to the field of data processing, in particular to a kind of data center operating temperature monitoring method, cabinet door detection method and monitoring system. Including: with predetermined time interval respectively obtain preset time window data center cold aisle refrigeration air temperature, cabinet air inlet temperature, cabinet ambient temperature;Utilize the refrigeration air temperature, the cabinet air inlet temperature, the cabinet ambient temperature calculates its pearson correlation coefficient and / or manhattan distance;According to the pearson correlation coefficient and / or manhattan distance judges whether the data center operating temperature is abnormal.The present application is based on the perception data center operating environment state, uses the way of data analysis to combine preset monitoring criterion and carries out temperature abnormality monitoring alarm.Meanwhile, through the logic processing mode, the abnormal criterion of monitoring alarm condition is enriched, the overall operation environment of cabinet can be analyzed, and the monitoring capability is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of data processing, in particular to a data center operating temperature monitoring method, a cabinet door opening and closing detection method and a monitoring system. BACKGROUND

[0002] With the rapid development of Internet business, more and more enterprises and institutions establish data centers to support the sustainable development of their businesses. Among them, the micro-module data center formed by placing row air conditioners and cabinets in a closed cold aisle for efficient management has gradually become a hot spot in data center construction. The micro-module data center adopts a closed cold aisle to improve the utilization rate of cold energy and the energy efficiency of the refrigeration system. The cold air sent by the precision air conditioner enters the closed cold aisle, is sucked by the cabinet fan to cool the load, and then is sent back to the precision air conditioner for circulation. In the above heat dissipation process, the monitoring and alarming of the environment temperature in the micro-module data center environment monitoring system is an important function, and the timeliness, accuracy and effectiveness of the alarm directly affect the quality of operation and maintenance monitoring and management. Therefore, real-time monitoring of the operating temperature of the data center is an important technical problem in order to enable staff to make timely treatment of abnormal temperature phenomena in the data center.

[0003] The monitoring system reads the temperature and humidity sensor data installed in the cabinet cold aisle from the intelligent device, and displays, analyzes and monitors the operating environment based on these data. At present, the abnormal temperature monitoring usually adopts the threshold overrun alarm method, that is, when the detection value of a certain temperature sensor of the cabinet exceeds the threshold of the pre-set monitoring temperature, the abnormal alarm module sends an alarm information, and when the detection value returns to the threshold, the abnormal alarm module suppresses the alarm information. This alarm method for monitoring temperature abnormalities is relatively simple and cannot reflect the operating state of the cabinet environment. It also needs to be analyzed by professional personnel on site to determine the alarm cause; for example, it is found that the cabinet door is opened or the cabinet sealing is reduced after on-site investigation. At the same time, the existing technology usually uses a door magnetic sensor to detect the opening and closing state of the cabinet door.

[0004] Another existing technology uses a method of comparing the pre-established temperature data range with the hysteresis value fluctuation range of the temperature data to determine whether the obtained temperature data is within the pre-established temperature data range. If not, an abnormal warning information is generated, otherwise no abnormal warning information is generated. When the temperature data is within the hysteresis value fluctuation range, the alarm state of the previous time is maintained to avoid repeated alarms caused by the temperature data jumping back and forth at the edge of the specified range. The above temperature monitoring method of the existing technology needs manual confirmation and judgment of the fault cause when the data exceeds the limit and an abnormal alarm is sent, which cannot effectively guide the operation and maintenance work. In addition, the temperature data of each monitoring point is determined separately, and the overall operating environment of the cabinet is not analyzed, which has limited ability to discover abnormalities. SUMMARY

[0005] To solve the above problems, the present application provides a data center operating temperature monitoring method, which utilizes air conditioning refrigeration supply air temperature, cabinet inlet air temperature and cabinet environment temperature sensors deployed in the data center cold aisle to perceive the data center operating environment state and monitor temperature abnormal state. Meanwhile, based on the monitoring method, a cabinet door opening and closing detection method is proposed, which can replace hardware such as door magnetic sensors to detect the cabinet door opening and closing state.

[0006] According to the first aspect, the present application provides a data center operating temperature monitoring method, comprising: acquiring refrigeration supply air temperature T supply , cabinet inlet air temperature T inlet and cabinet environment temperature T env of a data center cold aisle in a preset time window at a predetermined time interval; calculating Pearson correlation coefficient and / or Manhattan distance of the refrigeration supply air temperature T supply , the cabinet inlet air temperature T inlet and the cabinet environment temperature T env ; and judging whether the data center operating temperature is abnormal according to the Pearson correlation coefficient and / or Manhattan distance.

[0007] Preferably, the Pearson correlation coefficient and / or Manhattan distance of the refrigeration supply air temperature T supply , the cabinet inlet air temperature T inlet and the cabinet environment temperature T env is calculated by: calculating Pearson correlation coefficient P EI , P SE and P SI , wherein is the Pearson correlation coefficient of the refrigeration supply air temperature T supply and the cabinet inlet air temperature T inlet , is the Pearson correlation coefficient of the refrigeration supply air temperature T supply and the cabinet environment temperature T env , is the Pearson correlation coefficient of the cabinet environment temperature T env and the cabinet inlet air temperature T inlet ; and calculating Manhattan distance D EI , D SI and D SI , wherein is the Manhattan similarity of the refrigeration supply air temperature T supply and the cabinet inlet air temperature T inlet , is the Manhattan similarity of the refrigeration supply air temperature T supply and the cabinet environment temperature T env . Tcabinet is the temperature of the cabinet environment env Tcabinet is the temperature of the cabinet environment inlet Tcabinet is the temperature of the cabinet environment

[0008] Preferably, the judging whether the operation temperature of the data center is abnormal according to the Pearson correlation coefficient and / or the Manhattan distance comprises: judging whether D EI , D SI and D SI satisfy the following first relationship:

[0009]

[0010] If all satisfy, no abnormal alarm is generated; if any of the formulas does not satisfy, an abnormal alarm is generated.

[0011] Preferably, the judging whether the operation temperature of the data center is abnormal according to the Pearson correlation coefficient and / or the Manhattan distance comprises: judging whether P EI , P SE and P SI satisfy the following second relationship:

[0012]

[0013] If all satisfy, no abnormal alarm is generated; if any of the formulas does not satisfy, an abnormal alarm is generated.

[0014] Preferably, the judging whether the operation temperature of the data center is abnormal according to the Pearson correlation coefficient and / or the Manhattan distance comprises: judging whether P EI , P SE and P SI satisfy the following first relationship:

[0015]

[0016] and, D EI , D SI and D SI satisfy the following second relationship:

[0017]

[0018] If any of the first relationship does not satisfy and any of the second relationship does not satisfy, an abnormal alarm is generated; otherwise, no abnormal alarm is generated.

[0019] Preferably, the judging whether the operation temperature of the data center is abnormal according to the Pearson correlation coefficient and / or the Manhattan distance further comprises: utilizing the refrigeration supply air temperature T supply , the cabinet inlet air temperature T inlet , the cabinet environment temperature Tenv computing its fluctuation rate wherein is the refrigeration supply air temperature T supply fluctuation rate within a preset time window, is the cabinet ambient temperature T env fluctuation rate within a preset time window, is the cabinet inlet air temperature T inlet fluctuation rate within a preset time window.

[0020] Preferably, the judging whether the data center operating temperature is abnormal according to the Pearson correlation coefficient and / or Manhattan distance further comprises: if the abnormal alarm is suppressed; and if is not true, judging whether the data center operating temperature is abnormal according to the Pearson correlation coefficient and / or Manhattan distance.

[0021] Preferably, the predetermined time interval is 1 minute, and the length of the preset time window is 15 minutes.

[0022] According to a second aspect, the present application provides a data center cabinet door opening and closing detection method, comprising: executing the operating temperature monitoring method according to any one of the first aspect; if an abnormal alarm is generated, judging whether the following third relationship is satisfied based on P EI , P SI and D EI , D SI .

[0023]

[0024] If all are satisfied, it is determined that the cabinet door is in an open state.

[0025] According to a third aspect, the present application provides a monitoring system, comprising: a first temperature detector, a second temperature detector, a third temperature detector, a monitoring management platform; the first temperature detector is used to collect the refrigeration supply air temperature T supply of a data center cold aisle within a preset time; the second temperature detector is used to collect the cabinet inlet air temperature T inlet of the data center cold aisle within the preset time; the third temperature detector is used to collect the cabinet ambient temperature T env of the data center cold aisle within the preset time; and the monitoring management platform is used to execute the method according to any one of the first aspect and the second aspect.

[0026] According to a fourth aspect, the present application provides a storage medium, wherein the storage medium stores computer executable instructions, and when the computer executable instructions are loaded and executed by a processor, the steps of the method according to any one of the first aspect and the second aspect are implemented.

[0027] The operation temperature monitoring method of the present application combines sensors such as air conditioning refrigeration supply air temperature, cabinet inlet air temperature and cabinet environment temperature deployed inside the cold aisle of the data center to perceive the operation environment state of the data center, and uses data processing and analysis in combination with preset monitoring criteria to monitor and alarm temperature abnormalities. The monitoring criteria enrich the abnormal criteria of the monitoring alarm situation through logical processing, can analyze the overall operation environment of the cabinet, and improves the monitoring capability.

[0028] Meanwhile, the cabinet door opening and closing detection method based on the monitoring method in the present application can replace the use of hardware such as door magnetic sensors to detect the opening and closing state of the cabinet door, and realize the detection of the sealing property of the cabinet by data analysis, or serve as a supplementary means for verification. BRIEF DESCRIPTION OF DRAWINGS

[0029] Figure 1 is a deployment diagram of the monitoring system according to an embodiment of the present application in the cold aisle of the micro-module data center.

[0030] Figure 2 is a flowchart of the data center operation temperature monitoring method according to an embodiment of the present application.

[0031] Figure 3 is a flowchart of the data center cabinet door opening and closing detection method according to an embodiment of the present application.

[0032] Figure 4 is the effect of capturing the operation environment abnormal state by the embodiment of the present application when the load rate of the micro-module data center is 30%.

[0033] Figure 5 is the effect of capturing the operation environment abnormal state by the embodiment of the present application when the load rate of the micro-module data center is 60%.

[0034] Figure 6 is the effect of capturing the operation environment abnormal state by the embodiment of the present application when the load rate of the micro-module data center is 90%. DETAILED DESCRIPTION

[0035] The specific embodiments of the present application will be described in detail below, and it should be noted that the embodiments herein are only used for illustration and do not limit the present application. In the following description, a large number of specific details are set forth in order to provide a thorough understanding of the present application. However, it is obvious to those skilled in the art that the present application is not necessarily implemented by using these specific details. In other instances, well-known programs, materials or methods are not specifically described in order to avoid obscuring the present application.

[0036] The micro-module data center has a full-module architecture of refrigeration module, power supply and distribution module, intelligent monitoring and lighting, etc., and is highly integrated to realize rapid deployment and flexible expansion, and is suitable for different application environments. The cold and hot channels are fully enclosed, decoupled from the environment, close to the heat source, and the rack-type precision air conditioner and variable frequency fan are used, so that the air supply distance is shorter, and great advantages are achieved in saving cold loss and reducing energy consumption. The cold channel of the micro-module data center is a cooling measure to ensure the efficient operation of the micro-module data center. When normally operating, the temperature environment information of the cold channel of the data center is generally balanced and presents its own operation trend, and the relationship therebetween can be obtained by using a data analysis method to obtain corresponding data characteristics.

[0037] Figure 1 is a deployment diagram of a monitoring system according to an embodiment of the present application in a cold channel of a micro-module data center. In Figure 1 , three detection points are arranged in the cabinet to detect the operating environment temperature information in the cold channel. Among them, the first temperature detector 101 is arranged at the air conditioning refrigeration air supply area of the cold channel, and is used to collect the refrigeration air supply temperature T supply of the data center cold channel within a preset time; in some embodiments, the first temperature detector 101 can also be arranged near the air conditioning refrigeration air supply port or at a position near the air conditioning refrigeration air supply port in the cabinet; the second temperature detector 102 is arranged at the cabinet air inlet area of the cold channel, and is used to collect the cabinet air inlet temperature T inlet of the data center cold channel within a preset time; in some embodiments, the second temperature detector 102 is arranged at the cabinet air inlet or other positions that can collect data representing the cabinet air inlet temperature; the third temperature detector 103 is arranged at the top of the cold channel, and is used to collect the cabinet environment temperature T env of the data center cold channel within a preset time; in some embodiments, the third temperature detector 103 can also be arranged on the cabinet body or other positions that can collect data representing the cabinet environment temperature; the temperature information collected by the detector is sent to the monitoring management platform 104 through wired or wireless mode. In some embodiments, the first temperature detector 101, the second temperature detector 102 and the third temperature detector 103 are arranged from low to high around the cabinet. In some embodiments, they can also be arranged from high to low. The specific monitoring method of the monitoring management platform 104 will be described below in conjunction with Figure 2 .

[0038] Figure 2 is a flow chart of a data center operating temperature monitoring method according to an embodiment of the present application. Figure 2 The temperature monitoring method 200 shown can be applied to the monitoring system as shown in Figure 1 . Please refer to Figure 1 and Figure 2 together below.

[0039] In step S201, temperature environment information of the cold aisle of the data center in a preset time window is acquired at a predetermined time interval, the temperature environment information of the cold aisle comprising: refrigeration air supply temperature T supply , cabinet inlet air temperature T inlet , and cabinet environment temperature T env . In some embodiments, the predetermined time interval is preferably 1 minute, that is, the refrigeration air supply temperature, the cabinet inlet air temperature, and the cabinet environment temperature are collected once every 1 minute, and the preset time window is preferably 15 minutes, that is, the time series of the refrigeration air supply temperature, the cabinet inlet air temperature, and the cabinet environment temperature with a length of 15 minutes are acquired respectively. The time window is continuously sampled. According to other embodiments of the present application, the above sampling interval and the preset time length can be adjusted, for example, the sampling interval can be 30 seconds, and the preset time window length can be 10 minutes, etc.

[0040] In step S202, the Pearson correlation coefficient and / or the Manhattan distance of the refrigeration air supply temperature T supply , the cabinet inlet air temperature T inlet , and the cabinet environment temperature T env are calculated.

[0041] From the overall temperature field, the cold aisle operating temperature is mainly affected by the precision air conditioning refrigeration, the load heat dissipation, and the indoor environment temperature. In the case of normal refrigeration, the cabinet inlet air temperature T inlet of the cold aisle depends on the air conditioning refrigeration air supply temperature T supply , the temperature difference between the two is small, and the operation trend is consistent, and the cabinet environment temperature T env remains independent operation. Therefore, in some embodiments, the Pearson correlation coefficient is defined as an abnormal monitoring criterion.

[0042] The Pearson correlation coefficient is a statistical index reflecting the closeness of the correlation between variables. The correlation coefficient is calculated by the product difference method, which is also based on the deviation of two variables from their respective average values, and the correlation between variable X and variable Y is reflected by multiplying the two deviations. The Pearson correlation coefficient P XY of variable X and variable Y is:

[0043]

[0044] Wherein, Cov(X, Y) is the covariance of X and Y, Var[X] is the variance of X, Var[Y] is the variance of Y, n is the number of sampling points in the predetermined time window, x i is the value of X at the i-th sampling point, y i is the value of Y at the i-th sampling point, is the mean value of X in the predetermined time window, is the mean value of Y in the predetermined time window. P XY is a quantity that can characterize the closeness of the linear relationship between X and Y, and depicts the correlation degree of X and Y. P XY The absolute value of P is between 0 and 1. P XY The greater the absolute value of P is, the stronger the correlation degree between X and Y is, the more consistent the change trend of X and Y is, or the higher the mutual influence degree of X and Y is. P XY The greater the absolute value of P is, the stronger the correlation degree between X and Y is, the more consistent the change trend of X and Y is, or the higher the mutual influence degree of X and Y is. P XY The greater the absolute value of P is, the stronger the correlation degree between X and Y is, the more consistent the change trend of X and Y is, or the higher the mutual influence degree of X and Y is. P

[0045] The inventor finds that the Pearson correlation coefficient can be used to define an abnormal monitoring criterion, and applied to data center cold aisle temperature environment monitoring to reflect the running trend correlation degree between different variables. In the present application, P is the Pearson correlation coefficient of the refrigeration supply air temperature T supply and the cabinet inlet air temperature T inlet , is the Pearson correlation coefficient of the refrigeration supply air temperature T supply and the cabinet environment temperature T env , is the Pearson correlation coefficient of the cabinet environment temperature T env and the cabinet inlet air temperature T inlet . In the normal refrigeration condition, the data characteristics of the refrigeration supply air temperature T supply , the cabinet inlet air temperature T inlet and the cabinet environment temperature T env are shown as P SI >P EI , P SI >P SE ; otherwise, in the temperature abnormal condition caused by insufficient sealing of the cold aisle and the like, the cabinet environment temperature T env will affect the cabinet inlet air temperature T inlet and the refrigeration supply air temperature T supply , so that the correlation coefficients P EI and P SE become larger, and P SI becomes smaller. According to this correlation rule, one of the temperature conditions for normal operation of the data center can be set, that is, the Pearson correlation coefficient abnormal monitoring criterion is:

[0046]

[0047] Taking a preset time of 15 minutes and a sampling interval of 1 minute as an example, each refrigeration supply air temperature T supply , cabinet inlet air temperature T inlet and cabinet environment temperature T envThere are 15 sampling points in the time window, i.e. n = 15 in formula (1). Based on the refrigeration supply air temperature T supply , the cabinet inlet air temperature T inlet , and the cabinet ambient temperature T env , respectively, P EI , P SE , and P SI are calculated. If the values of P EI , P SE , and P SI satisfy the two relations of formula (2) above, it means that the temperature environment of the cold aisle of the data center is normal and no abnormality occurs; otherwise, it means that the temperature environment of the cold aisle of the data center may be abnormal.

[0048] In some embodiments, the Manhattan distance is used to define the abnormality monitoring criterion. The Manhattan similarity is defined based on the Manhattan distance. When the distance between two vectors is particularly small, it means that the two vectors have high similarity, and vice versa. Therefore, the distance metric is an index for measuring the similarity. The Manhattan distance is the sum of the projections of the line segment formed by two points on the fixed orthogonal coordinate system of the Euclidean space on the axes. In the plane, the Manhattan distance D XY of two variables X and Y is:

[0049]

[0050] where n is the number of sampling points in the preset time window, x i is the value of X at the i-th sampling point, and y i is the value of Y at the i-th sampling point.

[0051] The inventors have found that the Manhattan similarity can be used for monitoring the temperature environment of the cold aisle of a data center to reflect the closeness of the running trends between different variables. In the present application, D is the Manhattan similarity of the refrigeration supply air temperature T supply and the cabinet inlet air temperature T inlet , D is the Manhattan similarity of the refrigeration supply air temperature T supply and the cabinet ambient temperature T env , and D is the Manhattan similarity of the cabinet ambient temperature T env and the cabinet inlet air temperature T inlet . Under normal refrigeration, the cabinet inlet air temperature T inlet is greatly affected by the refrigeration supply air temperature T supply , the temperature difference between them is small, and the closeness is high, D SIsmaller. Conversely, in the case of abnormal refrigerant leakage in the cold aisle, the cabinet inlet temperature T inlet is affected, so that D EI becomes smaller, and D SI becomes larger. According to this correlation density law, one of the temperature conditions for normal operation of the data center can be set, that is, the Manhattan distance abnormality monitoring criterion is:

[0052]

[0053] Still taking 15 minutes as the preset time and 1 minute as the sampling interval, there are 15 sampling points in the time window of each refrigerant supply air temperature T supply , cabinet inlet temperature T inlet , and cabinet environment temperature T env , that is, n = 15 in formula (3). Based on the temperature values of the refrigerant supply air temperature T supply , cabinet inlet temperature T inlet , and cabinet environment temperature T env at the 15 sampling points, D EI , D SI , and D SI are calculated. If the calculated values satisfy the two relationship formulas of formula (4) above, it means that the temperature environment of the cold aisle of the data center is normal and no abnormality occurs; otherwise, it means that the temperature environment of the cold aisle of the data center may be abnormal.

[0054] In some embodiments, the volatility rate index is used to define the abnormality monitoring criterion. The volatility rate is a statistical concept, which is generally used to measure the degree of volatility of the price or investment return rate of the target asset. The volatility rate is the ratio of the difference between the maximum price and the minimum price in a certain period to the minimum price. For unit length time series data, the volatility rate V X of variable X is:

[0055]

[0056] Wherein, max(X) is the maximum value of X in the unit length time series data, and min(X) is the minimum value of X in the unit length time series data.

[0057] The inventors have found that the volatility rate can be used to reflect the volatility degree of the temperature at the detection point in the monitoring of the temperature environment of the cold aisle of the data center. In the present application, the volatility rate of the refrigerant supply air temperature T supply in the preset time window is V , the volatility rate of the cabinet environment temperature T env in the preset time window is V , and the volatility rate of the cabinet inlet temperature T inlet in the preset time window is V T inlet The fluctuation rate in the preset time window. In special cases, such as when the air conditioner cooling reaches equilibrium, the variance of the time series of each temperature in the cabinet is zero and presents a straight line trend, so that the Pearson correlation coefficient abnormal monitoring criterion is invalid. Therefore, the temperature condition for suppressing alarms in special cases of the data center is set, that is, the abnormal monitoring criterion of the fluctuation rate is:

[0058]

[0059] Still taking the preset time of 15 minutes and the sampling interval of 1 minute as an example, there are 15 sampling points in the time window of each refrigeration supply air temperature T supply , cabinet inlet air temperature T inlet , and cabinet environment temperature T env . Combined with formula (5), max(X) is the maximum value of X in the 15 data in the time series of unit length, and min(X) is the minimum value of X in the 15 data in the time series of unit length. When the above operating conditions are met, it indicates that the temperature environment of the cold aisle of the data center is in a special state, that is, the air conditioner cooling reaches equilibrium, so it is necessary to suppress abnormal alarms. That is, even if it is considered that a temperature anomaly may occur according to the Pearson correlation coefficient abnormal monitoring criterion at this time, no alarm is given, thereby avoiding false alarms; otherwise, it indicates that the temperature environment of the cold aisle of the data center may be abnormal, at which time further judgment needs to be made in combination with the Pearson correlation coefficient abnormal monitoring criterion or the Manhattan distance abnormal monitoring criterion. In some embodiments, to simplify the calculation in actual application, the difference between the maximum value and the minimum value of the collected temperature in the preset time can be directly calculated instead of the calculation of the fluctuation rate, so as to reduce the amount of calculation.

[0060] In step S203, it is determined whether the operating temperature of the data center is abnormal according to the Pearson correlation coefficient and / or the Manhattan distance. In some embodiments, it can be determined whether an alarm needs to be given based on one of the Pearson correlation coefficient abnormal monitoring criterion or the Manhattan distance abnormal monitoring criterion. That is, if one of the two conditions or neither of the two conditions in the Pearson correlation coefficient abnormal monitoring criterion is not established, or if one of the two conditions or neither of the two conditions in the Manhattan distance abnormal monitoring criterion is not established, it indicates that the temperature environment of the cold aisle of the data center may be abnormal, and an abnormal alarm is generated; otherwise, no alarm is generated.

[0061] Preferably, in some embodiments, the abnormality can also be determined based on both the Pearson correlation coefficient abnormal monitoring criterion and the Manhattan distance abnormal monitoring criterion, that is, when both the Pearson correlation coefficient abnormal monitoring criterion and the Manhattan distance abnormal monitoring criterion determine that there may be an abnormality, it is considered that the temperature environment of the cold aisle of the data center may be abnormal, and an abnormal alarm is generated; otherwise, no alarm is generated.

[0062] In a further preferred embodiment, the abnormality can also be judged based on both the Pearson correlation coefficient abnormality monitoring criterion and the volatility index abnormality monitoring criterion, that is, in the case where it is determined according to the Pearson correlation coefficient abnormality monitoring criterion that there may be an abnormality, the volatility index abnormality monitoring criterion is further judged, and if it is found that the volatility index abnormality monitoring criterion is not established at this time, it is indicated that the air conditioner refrigeration at this time reaches balance, the equipment temperature is normal, and no alarm is generated, that is, the Pearson correlation coefficient abnormality monitoring criterion is inhibited. Only in the case where the Pearson correlation coefficient abnormality monitoring criterion determines that there may be an abnormality and the volatility index abnormality monitoring criterion is not established, it is considered that the data center cold aisle temperature environment may appear abnormal, and an abnormality alarm is generated.

[0063] In a further preferred embodiment, the abnormality can also be judged based on the Pearson correlation coefficient abnormality monitoring criterion, the Manhattan distance abnormality monitoring criterion and the volatility index abnormality monitoring criterion, that is, in the case where it is judged based on both the Pearson correlation coefficient abnormality monitoring criterion and the Manhattan distance abnormality monitoring criterion that there may be an abnormality, the volatility index abnormality monitoring criterion is further judged, and if it is found that the volatility index abnormality monitoring criterion is established at this time, it is indicated that the air conditioner refrigeration at this time reaches balance, the equipment temperature is normal, and no alarm is generated, that is, the Pearson correlation coefficient abnormality monitoring criterion is inhibited. Only in the case where the Pearson correlation coefficient abnormality monitoring criterion and the Manhattan distance abnormality monitoring criterion both determine that there may be an abnormality and the volatility index abnormality monitoring criterion is not established, it is considered that the data center cold aisle temperature environment may appear abnormal, and an abnormality alarm is generated.

[0064] The present application utilizes the air conditioner supply air temperature, cabinet inlet temperature and environmental temperature sensors deployed in the data center cold aisle to perceive the data center operating environment state, and combines the Pearson correlation coefficient, Manhattan distance and volatility index and other mathematical statistical methods to monitor the cold aisle temperature balance state in a data analysis manner. The monitoring method is deployed in a digital monitoring and management platform, the temperature abnormality of the micro-module cold aisle operating environment can be captured from the level of the relationship between data, the cabinet refrigeration state is monitored, the monitoring criterion is combined in a logical processing manner to enrich the abnormality criterion of the monitoring alarm condition, the overall operating environment of the cabinet can be analyzed, the abnormal phenomenon that cannot be found by the conventional threshold determination method can be analyzed, the parameter-free configuration of the algorithm can adapt to the cold aisle temperature intelligent monitoring under different geographical positions and load conditions, the abnormal characteristics can be detected and timely alarm is generated, the corresponding processing is required, and the monitoring capability is improved.

[0065] Figure 3A flow chart of the cabinet door opening and closing detection method in a data center according to an embodiment of the present application. Figure 3 The cabinet door opening and closing detection method 300 in the data center can be applied to the monitoring system as shown in Figure 1 . Please refer to Figure 1 , Figure 2 and Figure 3 .

[0066] In step S301, the operating temperature monitoring method as shown in Figure 2 is performed.

[0067] In step S302, in the case that the operating temperature monitoring method generates an abnormal alarm, it is determined whether the cabinet door is in an open state based on the size relationship between P EI and P SI and the size relationship between D EI and D SI .

[0068] In some embodiments, the determination of whether the cabinet door is in an open state based on the size relationship between P EI and P SI and the size relationship between D EI and D SI includes judging whether the following third relationship is satisfied:

[0069]

[0070] If all the above conditions are satisfied, it indicates that the cabinet environment temperature T env and the cabinet inlet temperature T inlet have a high correlation and a close degree, and the cabinet door is in an open state.

[0071] In some embodiments, in the case that the third relationship is satisfied, it is further determined whether the fluctuation rate index abnormal monitoring criterion is satisfied. If it is found that the fluctuation rate index abnormal monitoring criterion is not satisfied, it is determined that the cabinet door is in an open state.

[0072] The embodiments of the present application can analyze the specific root cause of the temperature abnormality, such as the cold aisle baffle falling off or the cabinet door opening, after the temperature abnormality is monitored and an alarm is given in time, so as to achieve the purpose of applying the monitoring method of the embodiments of the present application to the detection of the cabinet door opening and closing state in the form of data analysis, and to achieve the effect of replacing the hardware detection by using a door magnetic sensor or as a supplementary means for verifying the hardware detection in the prior art.

[0073] Table 1 is a micro-module data center operating temperature characteristic analysis table, which reflects the data characteristics and correlation of the air conditioner supply air temperature, the cabinet inlet temperature and the cold aisle environment temperature.

[0074] ​Table 1. Micro-module data center operating temperature characteristic analysis table

[0075]

[0076] The following specific examples show the execution effect of the data center operating temperature monitoring method in the embodiment of the application.

[0077] Figure 4 The effect of capturing the abnormal state of the operating environment of the micro-module data center when the load rate is 30% in the embodiment of the application. From Figure 4 It can be concluded that when the low-load power section is normally operated, the cabinet cold aisle is maintained at a suitable temperature, the air supply temperature of the air conditioner refrigeration fluctuates obviously, and the rack inlet temperature affected by the air supply temperature also presents a synchronous fluctuation trend. Since the cold air in the aisle is inhaled by the cabinet to cool the load, and then circulated through the hot aisle, the cabinet environment temperature in the cold aisle remains a stable trend, which reflects the influence of the cabinet environment. The method of the application mainly analyzes data characteristics such as the air supply temperature of the air conditioner, the cabinet inlet temperature and the cold aisle environment temperature, and sends an alarm notification when the abnormal alarm criterion is met, such as Figure 4 From the 25th to the 43rd minute, it is the refrigeration abnormal alarm stage, and after the abnormal alarm criterion is invalid, the alarm is automatically restored.

[0078] Figure 5 The effect of capturing the abnormal state of the operating environment of the micro-module data center when the load rate is 60% in the embodiment of the application. From Figure 5 It can be concluded that when the high-load power section is operated in an abnormal environment, the air supply temperature of the air conditioner refrigeration fluctuates obviously, but the rack inlet temperature at this time is mainly affected by the cabinet environment temperature, resulting in different operating trends of the three. Due to the imbalance of the cold aisle, part of the cold air supplied by the air conditioner refrigeration leaks outside the cabinet, and part of it is inhaled by the cabinet to cool the load. In order to maintain the normal refrigeration supply of the load, the air conditioner is adjusted by frequency conversion to increase the refrigeration power consumption, thereby causing energy waste. For example, Figure 5 From the 33rd to the 83rd minute, it is the refrigeration abnormal alarm stage, and the air supply temperature of the air conditioner and the rack inlet temperature lose correlation and tightness, and the cold aisle refrigeration is imbalanced. After the cabinet door is closed again at the 83rd minute, the abnormal condition is invalid, and the alarm is automatically restored.

[0079] Figure 6 The effect of capturing the abnormal state of the operating environment of the micro-module data center when the load rate is 90% in the embodiment of the application. From Figure 6It can be concluded that when the high load power section is in normal operation, the air supply temperature of the air conditioner and the cabinet inlet temperature present a high correlation and tightness. When the cold aisle is unbalanced due to the opening of the cabinet door, the running environment is abnormal, and the data characteristics change to the cabinet environment temperature and the cabinet inlet temperature present a high correlation and tightness. At this time, the cabinet environment temperature and the cabinet inlet temperature are close to the indoor environment temperature. In order to maintain the normal heat dissipation demand of the load, the air conditioner adjusts the frequency to increase the refrigeration power consumption, thereby causing energy waste. For example Figure 6 From the 28th minute to the 93rd minute, it is the refrigeration abnormal alarm stage. After the cabinet door is closed again at the 93rd minute, the abnormal alarm criterion is invalid, and the alarm is automatically restored.

[0080] As can be seen from the above examples, the method in the embodiment of the present application can adapt to intelligent monitoring of the cold aisle temperature under different load conditions, analyze abnormal phenomena that are difficult to find, and perform accuracy statistics on the determination results during the test process, which illustrates the applicability and accuracy of the method. The test method of the above examples is to start from the time when the cabinet door is opened or closed, to perform abnormal determination and accuracy statistics on the temperature data in the time window, and to test different loads for 3 times, each time for 120 minutes. It can be compared from the test results in Table 2 that the time sequence determination of the 15-minute window can achieve better monitoring effect.

[0081] Table 2. Temperature monitoring experiment effect of data center running environment

[0082]

[0083] The embodiment of the present application includes the exemplary steps in sequence, but these steps do not have to be performed in the order shown. Performing these steps in different orders is within the scope of the present disclosure. In the spirit and scope of the embodiment of the present application, these steps can be added, replaced, changed in order and / or omitted as appropriate.

[0084] It should be noted that although the above describes each step in a specific order, it does not mean that each step must be performed in the above specific order. In fact, some of these steps can be performed concurrently, or even in a changed order, as long as the required functions can be achieved.

[0085] The present application can be a system, a method, and / or a computer program product. The computer program product can include a computer readable storage medium having computer readable program instructions loaded thereon, which are used to cause a processor to implement various aspects of the present application.

[0086] The computer readable storage medium can be a tangible device that can retain and store instructions for use by an instruction execution device. The computer readable storage medium can be, for example, but is not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. A non-exhaustive list of more specific examples of the computer readable storage medium includes the following: a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanically encoded device such as punch-cards or raised structures in a groove having instructions recorded thereon, and any suitable combination of the foregoing. A computer readable storage medium, as used herein, is not to be construed as being transitory signals per se, such as radio waves or other freely propagating electromagnetic waves.

[0087] Embodiments of the application have been described above with the aid of example implementations only. Numerous specific details have been set out to provide a thorough understanding of the embodiments. However, in certain circumstances, well-known methods and structures have not been described in order to avoid obscuring the concept. Only the preferred embodiments and / or variations of the embodiments have been described in detail, and obviously many modifications and changes can be made which fall within the scope of the embodiments. The above description is intended to be illustrative, and not restrictive. Many embodiments and applications other than the examples provided would be apparent with appreciation of the principles of the described embodiments to those of ordinary skill in the art. The scope of the technology should be determined with reference to the appended claims, along with the full scope of equivalents to which such claims are entitled.

[0088] BRIEF DESCRIPTION OF DRAWINGS

[0089] 101: first temperature detector

[0090] 102: second temperature detector

[0091] 103: third temperature detector

[0092] 104: monitoring management platform

[0093] 200: temperature monitoring method

[0094] S201, S202, S203: steps

[0095] 300: cabinet door opening and closing detection method

[0096] S301, S302: steps.

Claims

1. A method of monitoring operating temperatures in a data center, the method comprising: The method comprises: acquire the refrigeration supply air temperature of the cold aisle of the data center in a preset time window at a predetermined time interval , the cabinet inlet air temperature , the cabinet ambient temperature ; using the refrigerated supply air temperature the cabinet supply air temperature the cabinet ambient temperature calculate their Pearson correlation coefficient and Manhattan distance; the refrigeration supply air temperature the cabinet inlet air temperature the cabinet ambient temperature the fluctuation rate thereof , , wherein is the refrigeration supply air temperature in a preset time window, is the cabinet ambient temperature in a preset time window, is the cabinet inlet air temperature in a preset time window; According to the Pearson correlation coefficient, the Manhattan distance and the volatility, it is judged whether the operation temperature of the data center is abnormal, wherein if , the generation of an abnormal alarm is inhibited; and if is not true, according to the Pearson correlation coefficient and the Manhattan distance, it is judged whether the operation temperature of the data center is abnormal.

2. The method of claim 1, wherein, said refrigerated supply air temperature said cabinet supply air temperature said cabinet ambient temperature calculating a Pearson correlation coefficient and / or a Manhattan distance thereof, comprising: using Pearson's correlation coefficient , and wherein is the Pearson's correlation coefficient of the refrigerated supply air temperature and the cabinet inlet air temperature , is the Pearson's correlation coefficient of the refrigerated supply air temperature and the cabinet ambient temperature , is the Pearson's correlation coefficient of the cabinet ambient temperature and the cabinet inlet air temperature ; and Using Manhattan distance calculations , and where is the Manhattan similarity of the refrigerated supply air temperature to the cabinet intake air temperature , is the Manhattan similarity of the refrigerated supply air temperature to the cabinet ambient temperature , is the Manhattan similarity of the cabinet ambient temperature to the cabinet intake air temperature .

3. The method of claim 2, wherein, The judging whether the operation temperature of the data center is abnormal according to the Pearson correlation coefficient and / or Manhattan distance comprises: determining , and whether the following first relational expression is satisfied: ， If all the conditions are met, no abnormal alarm is generated; if any of the conditions is not met, an abnormal alarm is generated.

4. The method of claim 2, wherein, The judging whether the operation temperature of the data center is abnormal according to the Pearson correlation coefficient and / or Manhattan distance comprises: determining , and whether the following second relational expression is satisfied: ， If all the conditions are met, no abnormal alarm is generated; if any of the conditions is not met, an abnormal alarm is generated.

5. The method of claim 2, wherein, The judging whether the operation temperature of the data center is abnormal according to the Pearson correlation coefficient and / or Manhattan distance comprises: determining , and whether the following first relational expression is satisfied: ， and, determining , and whether the following second relational expression is satisfied: ， If any of the first conditions and any of the second conditions is not met, an abnormal alarm is generated; otherwise, no abnormal alarm is generated.

6. The method of claim 1, wherein, The predetermined time interval is 1 minute, and the length of the preset time window is 15 minutes.

7. A method for detecting the opening and closing of data center server rack doors, characterized in that, The method comprises: The method comprises: If an abnormal alarm is generated, it is determined whether the following third relational expression is satisfied based on , and , ​ ， If all the conditions are met, the cabinet door is determined to be in an open state.

8. A monitoring system, characterized by The method comprises: A first temperature detector, a second temperature detector, a third temperature detector, and a monitoring management platform. The first temperature detector is used to collect the cooling air supply temperature of the data center cold aisle within a preset time period. ; The second temperature detector is configured to collect the cabinet inlet air temperature of the cold aisle of the data center within the preset time. ; The third temperature detector is configured to collect the cabinet environment temperature of the cold aisle of the data center within the preset time. ; The monitoring management platform is configured to execute the method according to any one of claims 1 to 7.

9. A storage medium, characterized by The storage medium stores computer executable instructions, and when the computer executable instructions are loaded and executed by the processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images