Cabinet routing anomaly detection method and device based on current, medium and equipment

Abstract

The invention provides a current-based cabinet routing anomaly detection method and device, a medium and equipment, and belongs to the technical field of data processing. The method comprises the following steps: after a cabinet is powered on, acquiring each path of current data of each cabinet in a plurality of cabinets in a certain time period; a correlation value between each current data and each adjacent current data of the current data is calculated to form a correlation sequence of each current data, and the adjacent current data represents that the correlation sequence of each current data is greater than the correlation sequence of one path of current data of one cabinet. Different paths of current data of the same cabinet and each path of current data of a cabinet adjacent to the same cabinet are acquired; and based on the correlation sequence of the one or more current data, determining whether a power supply circuit wiring error of the plurality of cabinets exists or not. According to the invention, the power line fault detection efficiency and accuracy of the cabinet can be improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method, device, medium and equipment for detecting cabinet routing anomalies based on current. Background Art

[0002] In data center operations, servers are core components and critical assets. Their power supply stability is directly related to the overall operational efficiency and business continuity of the data center. To ensure stable server power, data centers generally use two or more power supply lines. For example, a two-line power supply model uses two power lines, A and B, for each cabinet. This ensures that even if one power line fails, the other line can still ensure normal power supply to the cabinet.

[0003] However, data centers have a large number of cabinets, and the physical layout of the header cabinet and several other cabinets is extremely compact, which poses a significant challenge to manual routing and wiring. In actual operation, wiring errors often occur. For example, a cabinet should be connected to both power supply A and B, but during operation, power supply A may be mistakenly connected to the power supply B of the adjacent cabinet, causing the cabinet to rely entirely on power supply B. If power supply B fails, the cabinet's power supply will be interrupted, causing the server to stop operating, severely impacting the data center's operations, and potentially causing data loss and service interruption.

[0004] Furthermore, during the long-term operation of a data center, various equipment failures cannot be ignored. Data collectors may fail, preventing accurate current data collection; loose wiring may cause single-circuit power outages. These failures not only impact server operation but also create significant challenges in troubleshooting. Manual troubleshooting of these types of failures often results in slow results, a high workload, and sometimes even difficulty identifying the fault immediately, further increasing the risk of data center operations.

[0005] Currently, data centers lack efficient and accurate fault monitoring methods, making it difficult to promptly detect and locate cabinet power supply failures. Existing monitoring methods mostly rely on manual inspections, which are not only inefficient but also prone to oversights. As data centers continue to expand in size and equipment, traditional monitoring methods are no longer able to meet the power supply stability and reliability requirements they demand. Therefore, developing a technology that can monitor cabinet current data in real time, quickly detect anomalies, and accurately locate problems is urgent. This is crucial for improving data center operations and management, and ensuring stable business operations. Summary of the Invention

[0006] Based on this, the purpose of this application is to provide a current-based cabinet routing anomaly detection method, device, medium and equipment to improve the timeliness of anomaly identification of cabinets in a data center and the accuracy of anomaly positioning.

[0007] In a first aspect of the present application, a method for detecting cabinet routing anomalies based on current is provided, the method comprising:

[0008] After the cabinet is powered on, obtain the current data of each circuit in each cabinet in a certain time period;

[0009] Calculating the correlation value between each current data and each adjacent current data of the current data to form a correlation sequence for each current data, wherein the adjacent current data represents, with respect to one current data of one cabinet, different current data in the same cabinet and current data of cabinets adjacent to the same cabinet;

[0010] It is determined whether there is a wiring error of power lines of the cabinets in the plurality of cabinets based on a correlation sequence of the one or more current data.

[0011] Optionally, each cabinet includes two channels of current data, and the adjacent current data of one channel of current data of any cabinet include the other channel of current data of the cabinet and the two channels of current data of two cabinets adjacent to the cabinet. The correlation sequence of each current data includes the correlation value between the current data and each adjacent current data of the current data.

[0012] Optionally, the method of determining whether there is a power line wiring error in the multiple cabinets based on the correlation sequence of one or more current data includes: when the correlation value between two current data of the same cabinet is a smaller value in the correlation sequence of any current data of the same cabinet, determining that the power line wiring of the cabinet is incorrect.

[0013] Optionally, the determining whether there is a power line wiring error in the cabinet among the multiple cabinets based on the correlation sequence of one or more current data includes: selecting first target current data and second target current data from the multiple current data; determining whether there is a power line wiring error in the target cabinet based on a first correlation sequence of the first target current data and a second correlation sequence of the second target current data, the first target current data and the second target current data being two current data of the target cabinet, respectively.

[0014] Optionally, determining whether a power line wiring error exists in the target cabinet based on a first correlation sequence of the first target current data and a second correlation sequence of the second target current data includes:

[0015] determining whether a first target correlation value in the first correlation sequence is a larger value in the first correlation sequence, the first target correlation value being a correlation value of the second target current data relative to the first target current data;

[0016] determining whether a second target correlation value in the second correlation sequence is a larger value in the second correlation sequence, the second target correlation value being a correlation value of the first target current data relative to the second target current data;

[0017] When both values ​​are larger, it is determined that the target cabinet power line wiring is correct;

[0018] When either value is not a larger value, it is determined that the target cabinet power supply circuit is incorrectly wired.

[0019] Optionally, each cabinet includes two channels of current data; and determining whether there is a power line wiring error in the cabinets based on the correlation sequence includes:

[0020] The object corresponding to the maximum correlation value in each correlation sequence is regarded as the most relevant object;

[0021] If the most relevant object of the two-way current data of the target cabinet is the current data of the same way of the adjacent cabinet, and the most relevant object of the two-way current data of the adjacent cabinet is also the current data of the same way of the target cabinet, it is determined that the different power supplies of the target cabinet and the adjacent cabinet are connected incorrectly;

[0022] If the most relevant object of the two-way current data of the target cabinet is the different-way current data of the adjacent cabinet, and the most relevant object of the two-way current data of the adjacent cabinet is also the different-way current data of the target cabinet, it is determined that the same-way power supply of the target cabinet and the adjacent cabinet is connected incorrectly.

[0023] Optionally, calculate the current data I of the i-way current of cabinet m according to the following formula: m_i The j-way current data I of cabinet n n_j The correlation value C m_i / n_j :

[0024]

[0025] K represents the number of data points in the current data, m_i k Indicates current data I m_i The value of the kth data point in n_j k Indicates current data I n_j The value of the kth data point in .

[0026] Optionally, after obtaining the current data of each channel of each cabinet in the plurality of cabinets within a certain time period, the method further includes:

[0027] Identify whether there are persistent missing values ​​or zero values ​​in the current data of each channel of each cabinet. If so, determine that the corresponding cabinet has a power outage or data collection failure; and / or

[0028] Identify whether there are multiple consecutive data points in each current data of each cabinet that are all smaller than a first threshold value. If so, determine that the corresponding cabinet is not configured with a computing device or is in a test condition; and / or

[0029] Identify whether the maximum fluctuation value between data points in each current data of each cabinet is less than the second threshold and not 0 at the same time. If so, determine that the corresponding cabinet is not configured with computing equipment or is in a test condition.

[0030] In a third aspect of the present application, a computer-readable storage medium is provided, on which executable instructions are stored. When the executable instructions are executed by a processor, the processor executes the method described in any embodiment of the present application.

[0031] In a fourth aspect of the present application, an electronic device is provided, comprising: one or more processors; and a memory for storing one or more programs, which, when executed by the one or more processors, enables the one or more processors to execute the method described in any one of the embodiments of the present application.

[0032] The current-based cabinet routing anomaly detection method, device, medium, and equipment in the present application calculate the correlation value between current data and its adjacent current data to form a correlation sequence. By analyzing the correlation sequence, it is possible to locate which specific cabinet has a wiring error, thereby facilitating relevant staff to directly conduct line inspections at the cabinet corresponding to the wiring error, thereby improving the efficiency and accuracy of fault detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope of the present application.

[0034] Figure 1 1 is a flow chart of a method for detecting abnormality in cabinet routing based on current in one embodiment;

[0035] Figure 2 This is a schematic diagram of power line wiring of a cabinet in one embodiment;

[0036] Figure 3A schematic diagram of a process for determining whether a target cabinet has a power line wiring error based on a first correlation sequence of first target current data and a second correlation sequence of second target current data in one embodiment;

[0037] Figure 4 1 is a flow chart of a process for determining the type of wiring error in one embodiment;

[0038] Figure 5 1 is a flow chart of a method for detecting abnormality in cabinet routing based on current in another embodiment;

[0039] Figure 6 1 is a schematic structural diagram of a current-based cabinet routing anomaly detection device in one embodiment;

[0040] Figure 7 FIG. 4 is a structural block diagram of an electronic device in an embodiment. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solutions and advantages of this application more clearly understood, the present application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0042] All terms (including technical and scientific terms) used in this application have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0043] For example, the terms "first," "second," etc. used in this application may be used herein to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish a first element from another element.

[0044] For example, the terms "include", "comprising", etc. used in this application indicate the existence of features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0045] In one embodiment, Figure 1 As shown, a current-based cabinet routing anomaly detection method is provided, the method comprising:

[0046] Step 110 : After the cabinet is powered on, obtain current data of each channel of each cabinet in a certain time period.

[0047] In this embodiment, the power distribution and power-on status information of each cabinet is obtained from the data center monitoring platform, and current data is collected for the cabinets that have been powered on and powered on. The same cabinet includes multiple power supplies, and the number of power supplies may include two or more. Taking the two-way power supply mode as an example, the same cabinet includes power supply A and power supply B. The cabinet represents a device that integrates, stores or manages one or more computer hardware (such as the above-mentioned graphics processing unit GPU, data processing unit DPU, and central processing unit CPU). The cabinet is arranged in the computer room of the data center, which can specifically be an IDC computer room where multiple cabinets are arranged.

[0048] Each row of cabinets is equipped with a row head cabinet. The main function of the row head cabinet is to distribute power to the cabinets and monitor the power / current and other power data distributed to the cabinets. Specifically, in the two-way (A and B) power supply mode, the row head cabinets can include the A-way row head cabinet and the B-way row head cabinet. Figure 2 As shown, the air switch QF35 of the A-line terminal cabinet 9A and the air switch QF35 of the B-line terminal cabinet 10B are respectively connected to the two PDU strips of cabinet F16, providing A-line power and B-line power for cabinet F16; the air switch QF36 of the A-line terminal cabinet 9A and the air switch QF36 of the B-line terminal cabinet 10B are respectively connected to the two PDU strips of cabinet F15, providing A-line power and B-line power for cabinet F15.

[0049] Under normal circumstances, the A and B terminal cabinets jointly distribute power to the cabinets. If either the A or B terminal cabinet fails, the other terminal cabinet can independently complete power distribution. By monitoring the power and current distributed to each cabinet by the terminal cabinet, the power and current of each cabinet can be obtained. Specifically, the A and B terminal cabinets of the terminal cabinet can monitor and collect the power and current data of both the A and B terminals distributed to each cabinet.

[0050] The time period can be a set time period with a certain duration, which can be a pre-set fixed duration or any suitable duration customized by the user. The selected time period should be able to reflect the current changes under the normal business load of the data center. For example, any suitable duration such as 10 minutes, 15 minutes, 30 minutes, one hour, etc. during the peak business period of the data center is selected as a time period, and the current data of each cabinet in the time period is obtained. It can be understood that the electronic device can collect current data at a preset frequency, such as once per minute, once every 5 seconds, once every 10 seconds, etc.

[0051] The current data of each power supply of each cabinet collected in a time period has multiple current data. Each current data of the cabinet represents the current data of each power supply of the cabinet. The current data of channel A IX_A Indicates the current data of the A power supply of cabinet X, the current data of the B power supply I X_B This represents the current data for power supply channel B of cabinet X. The two current data points for a cabinet represent the current data for channel A and channel B of cabinet X. A data point in the current data can be a specific current value, or a normalized value generated by processing actual measured current values ​​according to preset processing rules. The current data for a time period can be a sequence of current values ​​or normalized values ​​at each moment in that time period. For example, a current data set for a certain time period might contain 300 data points, such as I = [10.2A, 10.3A, 10.1A, ..., 10.5A].

[0052] Step 120 : Calculate the correlation value between each current data and each adjacent current data of the current data to form a correlation sequence for each current data.

[0053] In this embodiment, adjacent current data is a relative concept of current data, which is adjacent current data relative to a certain current data. Taking one of the current data as the target current data, the adjacent current data of the target current data represents other current data that are closely related to the target current data. This close relationship is reflected in the proximity of the positions of the cabinets corresponding to the two (the target current data and its adjacent current data) or the proximity of the number of power supply paths, such as the cabinets being adjacent to each other, or the two cabinets being within a preset distance range.

[0054] In one embodiment, when close association is used to indicate that two cabinets are adjacent, adjacent current data represents different current data of the same cabinet and current data of cabinets adjacent to the same cabinet relative to one current data of one cabinet.

[0055] For example, if a current data point is used as the current data to be analyzed, the adjacent current data for the current data to be analyzed includes the current data for the other power sources of the cabinet to be analyzed (i.e., the power source to be analyzed other than the current data point). It also includes the current data for each cabinet adjacent to the cabinet to be analyzed. Adjacent refers to the cabinets being physically adjacent.

[0056] The cabinets in the data center can be placed in the computer room and arranged in order. For example, if there are N cabinets in the computer room of the data center, and the N cabinets are arranged in order, then cabinet X represents the cabinet at the Xth position (N≥X≥1). In the two-way (A and B) power supply mode, I X_A Indicates the current data of channel A of cabinet X; I X_B Indicates the current data of channel B in cabinet X.

[0057] In one embodiment, each cabinet includes two channels of current data, and the adjacent current data of one channel of current data of any cabinet include the other channel of current data of the cabinet and the two channels of current data of two cabinets adjacent to the cabinet. The correlation sequence of each current data includes the correlation value between the current data and each adjacent current data of the current data.

[0058] For example, the current data of channel A and channel B are two channels. The current data of channel A of cabinet X is I X_A The adjacent current data includes: Cabinet X B current data I X_B , the current data I of the A path of the cabinet X-1 adjacent to the cabinet X X-1_A 、Channel B current data I of cabinet X-1 X-1_B , the current data I of the A path of the cabinet X+1 adjacent to the cabinet X X+1_A , Cabinet X+1 B-channel current data I X+1_B These adjacent current data form I X_A The adjacent current data sequence {I X_B , I X-1_A , I X-1_B , I X+1_A , I X+1_B}. It is understandable that the current data of cabinet X, channel B, is I X_B The adjacent current data sequence is {I X_A , I X-1_A , I X-1_B , I X+1_A , I X+1_B}.

[0059] Calculate the correlation value between each adjacent current data of the current data and the current data to obtain the correlation sequence of the current data. X_A For example, specifically, calculate the current data I X_A Each adjacent current data I · With the current data I X_A The correlation value C X_A / · , get the current data I X_A The correlation sequence C X_A . “I · " can be the adjacent current data sequence {I X_B , I X-1_A , I X-1_B , I X+1_A , I X+1_B}. Among them, the current data I X_A The correlation sequence C X_A Includes: C X_A / X_B 、C X_A / X-1_A 、CX_A / X-1_B , C X_A / X+1_A , C X_A / X+1_B or more of the above, C X_A / X_B represents the current data I X_B relative to the current data I X_A The correlation value. It can be understood that the current data I X_B The correlation sequence C of[[ID=...]] X_B includes: C X_B / X_A , C X_B / X-1_A , C X_B / X-1_B , C X_B / X+1_A , C X_B / X+1_B or more of the above.

[0060] It can be understood that when X = 1, the correlation sequence C of the current data I 1_A includes {C 1_A , C 1_A / 1_B , C 1_A / 2_A}, when X = N, the correlation sequence C of the current data I 1_A / 2_B includes {C N_A , C N_A , C N_A / N_B , C N_A / N-1_A , C N_A / N-1_B}, when 1 < X < N, the correlation sequence C of the current data I X_A includes {C X_A , C X_A / X_B , C X_A / X-1_A , C X_A / X-1_B , C X_A / X+1_A , C X_A / X+1_B}.

[0061] Among them, the calculation of the correlation value between two current data can be carried out according to a pre-set correlation calculation model. The correlation calculation model can be a combination of one or more of the Pearson correlation coefficient, Spearman rank correlation coefficient, and cosine similarity correlation coefficient calculation algorithms.

[0062] In one embodiment, the correlation value C m_i between the i-th current data I n_j of cabinet m and the j-th current data I m_i / n_j of cabinet n is calculated according to the following formula:

[0063]

[0064] where K represents the number of data points in the current data, m_i k represents the value of the k-th data point in the current data I m_i (i.e., the i-th current data of cabinet m (the cabinet at the m-th position)), and n_j k represents the current data In_j (i.e. the value of the kth data point in the jth current data of cabinet n (cabinet at the nth position)). For example, if there are 300 data points in a current data I, then K = 300, and the current data I can be m_i and current data I n_j Substitute the corresponding data points into the above formula to calculate C m_i / n_j It can be understood that the above formula can be used to calculate the C X_A / X_B 、C X_A / X-1_A 、C X_A / X-1_B 、C X_A / X+1_A 、C X_A / X+1_B .

[0065] For example, I m_i =[10,12,15,13,16], I n_j =[11,13,16,14,17], then K=5, and the above formula can be used to calculate The final calculation of C is obtained by this formula m_i / n_j =1.

[0066] Step 130 : Determine whether there is a wiring error in the power lines of the cabinets in the plurality of cabinets based on the correlation sequence of the one or more current data.

[0067] In this embodiment, each correlation value in the correlation sequence reflects the correlation between the current data to be analyzed and one of its adjacent current data. By horizontally comparing the correlation values ​​in the correlation sequence, a ranking of the correlations between the current data to be analyzed and its adjacent current data can be determined, thereby determining whether the cabinet corresponding to the current data has a power line wiring error.

[0068] Generally speaking, in the correlation sequence of the same current data, the correlation values ​​of different current data of the same cabinet are larger in the correlation sequence, while the correlation values ​​of different current data of different cabinets are smaller in the correlation sequence. If the correlation values ​​in the analyzed correlation sequence do not meet this feature, it means that the cabinet corresponding to the correlation sequence has a power line wiring error.

[0069] Furthermore, verification can be performed by analyzing the ranking of correlation values ​​in multiple correlation sequences in the correlation sequences, so as to further improve the accuracy of determining whether there is a power line wiring error in the corresponding cabinet.

[0070] That is, for the correlation sequence C X_A The larger of the correlation values ​​is usually the correlation value C X_A / X_B , the smaller one should usually be the correlation value C X_A / X-1_B、C X_A / X+1_B For example, the correlation sequence C X_A :{C X_A / X_B 、C X_A / X-1_A 、C X_A / X-1_B 、C X_A / X+1_A 、C X_A / X+1_B}={0.95, 0.88, 0.70, 0.75, 0.66}. Among them, C X_A / X_B The value is the largest, and C X_A / X-1_B 、C X_A / X+1_B The value is the smallest. If the correlation value in a correlation sequence does not meet this feature, but C X_A / X_B The value is small, while C X_A / X-1_B 、C X_A / X+1_B If the value is large, it means that there is a power line wiring error in cabinet X.

[0071] Specifically, when the correlation value between two current data of the same cabinet is a smaller value in the correlation sequence of any current data of the same cabinet, it is determined that the power line wiring of the cabinet is incorrect.

[0072] For example, when C X_A / X_B In the correlation sequence C X_A When the value is smaller, it is determined that the power line wiring of cabinet X is incorrect.

[0073] In one embodiment, further, when the correlation sequence C X_A When the correlation value in does not meet the above characteristics (that is, the larger value is not C X_A / X_B , the smaller value is not C X_A / X-1_B 、C X_A / X+1_B ), in order to improve the accuracy of determining the power line wiring error, the correlation sequence C X_B Does it also not meet the above characteristics? If the correlation sequence C X_B It does not meet this characteristic (that is, the larger value is not C X_B / X_A , the smaller value is not C X_B / X-1_A 、C X_B / X+1_A ), it means that the power line wiring of cabinet X is incorrect.

[0074] That is, when C X_A / X_B In the correlation sequence C X_A is a smaller value, and C X_B / X_A In the correlation sequence C X_B When the value is smaller, it is determined that the power line wiring of cabinet X is incorrect.

[0075] The larger value can be the first a values ​​in the descending order, and the smaller value can be the last b values. a and b can be determined based on the number c of correlation values ​​in the correlation sequence, for example, a < (c / 2), b < (c / 2). Taking c = 5 as an example, the larger value can be the largest or second largest value, and the smaller value can be the smallest or second smallest value.

[0076] Furthermore, the larger value includes the maximum value in the correlation sequence, and a non-maximum value whose absolute difference from the maximum value is less than a difference threshold. The difference threshold can be a preset fixed value, or a value adaptively determined based on the magnitude of each correlation value in the correlation sequence. For example, it can be a value adaptively determined based on the correlation values ​​in all correlation sequences obtained in the current time period or in the most recent time periods.

[0077] Among them, power line wiring errors include misconnections between different power supplies and misconnections between power supplies on the same line. Misconnections between different power supplies usually occur when different power supplies in the same cabinet are misconnected, for example, the power interface of cabinet X that should be connected to power supply A is connected to power supply B; it can also occur when different power supplies in different cabinets are misconnected, for example, the power interface of cabinet X that should be connected to power supply A is connected to power supply B of cabinet X+1 (or cabinet X-1, etc.). Misconnections between power supplies on the same line usually occur when the same power supplies in different cabinets are misconnected, for example, the power interface of cabinet X that should be connected to power supply A of cabinet X is connected to power supply A of cabinet X+1 (or cabinet X-1, etc.).

[0078] The current-based cabinet routing anomaly detection method in this application calculates the correlation value between current data and its adjacent current data to form a correlation sequence. By analyzing the correlation sequence, it is possible to locate which specific cabinet has a wiring error, thereby facilitating relevant staff to directly conduct line inspections at the cabinet corresponding to the wiring error, thereby improving the efficiency and accuracy of fault detection.

[0079] In one embodiment, step 130 includes: selecting first target current data and second target current data from a plurality of current data; and determining whether a power line wiring error exists in the target cabinet based on a first correlation sequence of the first target current data and a second correlation sequence of the second target current data.

[0080] In this embodiment, multiple correlation sequences may be selected for verification, and the magnitude of each correlation value in the selected correlation sequence may be analyzed to determine whether a wiring error exists in the corresponding cabinet.

[0081] Specifically, the first target current data and the second target current data are two current data of the target cabinet, for example, the first target current data and the second target current data are current data of channel A and current data of channel B of the target cabinet, respectively.

[0082] By analyzing the magnitude relationship between the correlation values ​​in the first correlation sequence and the second correlation sequence, it is determined whether the target cabinet has a wiring error. For example, the larger and smaller values ​​in the two correlation sequences can be analyzed to see if they meet the aforementioned characteristics. If neither meets the aforementioned characteristics, the target cabinet is determined to have a wiring error.

[0083] In one embodiment, Figure 3 As shown, determining whether a power line wiring error exists in a target cabinet based on a first correlation sequence of first target current data and a second correlation sequence of second target current data includes:

[0084] Step 310: Determine whether a first target correlation value in the first correlation sequence is a larger value in the first correlation sequence.

[0085] Step 320: Determine whether the second target correlation value in the second correlation sequence is a larger value in the second correlation sequence.

[0086] Step 330: When both values ​​are larger, it is determined that the target cabinet power supply circuit is correctly wired.

[0087] Step 340: When either of the two is not a larger value, it is determined that the target cabinet power supply circuit is incorrectly wired.

[0088] In this embodiment, the first target correlation value is the correlation value of the second target current data relative to the first target current data, and the second target correlation value is the correlation value of the first target current data relative to the second target current data. Taking cabinet X as an example, its first target current data is I X_A , the first correlation sequence is C X_A , the second target current data is I X_B , the second correlation sequence is C X_B The first target correlation value is C X_A / X_B , the second target correlation value is C X_B / X_A .

[0089] By selecting two correlation sequences corresponding to the same cabinet, it is determined whether the selected cabinet has a wiring error based on the size relationship of the correlation values ​​in the two correlation sequences and the size order of the correlation values ​​in their respective correlation sequences, thereby improving the accuracy of wiring error identification.

[0090] In one embodiment, the current-based cabinet routing anomaly detection method further includes: when there is a wiring error in the power supply line of the cabinet, further determining a specific wiring error type.

[0091] like Figure 4 As shown in the figure, the process of determining the wiring error type includes:

[0092] Step 410: The object corresponding to the maximum correlation value in the correlation sequence of each current data is taken as the most correlated object.

[0093] For example, each cabinet includes two current data, and the two current data can be the above-mentioned current data of channel A and channel B. If the target cabinet is cabinet X, the correlation sequence C X_A :{C X_A / X_B 、C X_A / X-1_A 、C X_A / X-1_B 、C X_A / X+1_A 、C X_A / X+1_B}={0.95, 0.88, 0.70, 0.75, 0.66}, then the correlation sequence C of the current data of channel A of cabinet X is X_A The maximum correlation value C X_A / X_B =0.95, the maximum correlation value C X_A / X_B The corresponding object is X_B, that is, the most relevant object of the current data of channel A of cabinet X is the current data of channel B of cabinet X.

[0094] In step 420, if the most relevant object of the two-way current data of the target cabinet is the same-way current data of the adjacent cabinet, and the most relevant object of the two-way current data of the adjacent cabinet is also the same-way current data of the target cabinet, then it is determined that the wiring error type is that the target cabinet and its adjacent cabinet are connected to different power supplies incorrectly.

[0095] If the most relevant object of the current data of channel A of the target cabinet X is the current data of channel A of cabinet X+1, it can be expressed as: Mc["X_A"] == "(X+1)_A".

[0096] When the most relevant object of any one of the two current data of a cabinet is the corresponding same-channel current data of its adjacent cabinet, it means that the most relevant object of the two current data of the cabinet is the corresponding same-channel current data of the adjacent cabinet.

[0097] Specifically, when Mc["X_A"] == "(X+1)_A" and Mc["X_B"] == "(X+1)_B" both hold true (i.e., the most relevant object for the current data of channel A of target cabinet X is the current data of channel A of cabinet X+1, and the most relevant object for the current data of channel B of target cabinet X is the current data of channel B of cabinet X+1), it means that the most relevant objects for the two current data of cabinet X are the current data of the same channels in its adjacent cabinet X+1. When both Mc["(X+1)_A"] == "X_A" and Mc["(X+1)_B"] == "X_B" hold true (i.e., the most relevant object for the current data on channel A of target cabinet X+1 is the current data on channel A of cabinet X, and the most relevant object for the current data on channel B of target cabinet X+1 is the current data on channel B of cabinet X), then the most relevant objects for the two current data on the adjacent cabinet X+1 are also the current data on the same channel of target cabinet X. At this point, it can be determined that different power supplies of cabinet X and its adjacent cabinet X+1 are incorrectly connected. For example, the power interface of cabinet X that should be connected to power supply A is connected to power supply B of cabinet X+1, and / or the power interface of cabinet X that should be connected to power supply B is connected to power supply A of cabinet X+1.

[0098] In step 430, if the most relevant object of the two-way current data of the target cabinet is the different-way current data of the adjacent cabinet, and the most relevant object of the two-way current data of the adjacent cabinet is also the different-way current data of the target cabinet, then it is determined that the wiring error type is that the target cabinet and its adjacent cabinet have the same power supply connected incorrectly.

[0099] Similarly, when M_c["X_A"] == "(X+1)_B", M_c["X_B"] == "(X+1)_A", M_c["(X+1)_A"] == "X_B", and M_c["(X+1)_B"] == "X_A" all hold true, it indicates that the wiring error type is that the power supplies of cabinet X and the adjacent cabinet X+1 are incorrectly connected. For example, the power port of cabinet X that should be connected to cabinet X's power supply A is connected to cabinet X+1's power supply A, and / or the power port of cabinet X that should be connected to cabinet X's power supply B is connected to cabinet X+1's power supply B.

[0100] In one embodiment, Figure 5 As shown, another current-based cabinet routing anomaly detection method is provided, the method comprising:

[0101] Step 510: After the cabinet is powered on, obtain current data of each channel of each cabinet in a certain time period.

[0102] After all cabinets in the data center have been powered on and running stably for a certain period of time, they can be divided into time periods according to preset unit lengths. The values ​​collected at a preset frequency during these time periods are processed to form corresponding data points in the current data. Based on this, the current data for each channel in each cabinet can be obtained. A current data point corresponds to the current value of one channel in a cabinet during a time period.

[0103] Step 520 , identifying whether there are persistent missing values ​​or 0 values ​​in each current data of each cabinet. If so, it is determined that the corresponding target cabinet has a power outage or data collection failure.

[0104] For example, the power supply in cabinet X includes channels A and B. For channel A power supply in cabinet X, the corresponding channel A current data is recorded as I X_A , the B-way current data corresponding to the B-way power supply of cabinet X is I X_B If there are persistent missing values ​​or 0 values ​​in the two current data collected for cabinet X, it is determined that cabinet X has a power outage or data collection failure.

[0105] Specifically, when I X_A , I X_B If there are persistent missing values ​​or 0 values, and the duration of the missing values ​​or 0 values ​​exceeds the preset first duration threshold, it means that cabinet X has been powered off for a long time when the system is powered on. The possible cause is an actual power outage on site, or a fault in the data acquisition module or transmission equipment that lasts for a long time.

[0106] At this time, the electronic equipment may generate an alarm message: Cabinet X may have experienced a long-term actual power outage on site, or the data acquisition module or transmission equipment may have failed. Please check!

[0107] When I X_A , I X_B If the duration of missing or zero values ​​exceeds the preset second duration threshold but is less than the first duration threshold, it indicates that cabinet X may have experienced a momentary power outage or a failure in the data acquisition module or transmission equipment. The first and second duration thresholds can be any appropriate values ​​set based on actual circumstances.

[0108] At this time, the electronic equipment may generate an alarm message: Cabinet X may have a momentary on-site power outage, or the data acquisition module or transmission equipment may have failed. Please check!

[0109] Step 530 , identifying whether there are multiple consecutive data points in each current data of each cabinet that are all smaller than a first threshold value. If so, it is determined that the corresponding target cabinet is not configured with a computing device or is in a test condition.

[0110] Optionally, it is possible to identify whether the values ​​of all data points in each current data of each cabinet are less than the first threshold at the same time. If so, it is determined that the cabinet is not equipped with computing equipment or is in a test condition. For example, when I X_A The values ​​in are neither 0 nor empty, and are all less than the first threshold. At the same time, I X_B If the values ​​in are not 0 or empty and are all less than the first threshold, it means that cabinet X may be configured with only network devices but no computing devices, or it is a test condition and has no business analysis value.

[0111] The first threshold can be set based on the actual value of the current data in the data center. The number of consecutive data points can be determined based on the current data collection duration and frequency. For example, if a data point is required to be less than the first threshold for a unit time duration, the number of consecutive data points can be determined based on the unit time duration and the collection frequency.

[0112] Step 540 , identifying whether the maximum fluctuation value between data points in each current data of each cabinet is less than the second threshold and not 0 at the same time; if so, determining that the corresponding target cabinet is not configured with computing equipment or is in a test condition.

[0113] The sizes of the first threshold and the second threshold can refer to the server power and current. For example, the power of a server is about 200W, the current is about 1A, and the load change during the business cycle is greater than 0.2A. In this case, the first threshold can be set to 1A and the second threshold can be set to 0.2A.

[0114] The maximum fluctuation value can be specifically the absolute value of the difference between the maximum value and the minimum value in the corresponding current data. X_A The difference between the maximum and minimum values ​​is less than 0.2A. X_B The difference between the maximum and minimum values ​​is also less than 0.2A.

[0115] When I X_A and I X_B If the maximum fluctuation values ​​of are all greater than 0 and less than the second threshold, it means that there is fluctuation, but the abnormal situation may be that the cabinet is only configured with network equipment but not computing equipment, or the test condition has no business analysis value.

[0116] At this point, the electronic equipment may generate an alarm message: Cabinet X may only be configured with network equipment and no computing equipment, or the cabinet is in a test condition and has no business analysis value. Please verify!

[0117] In one embodiment, when the maximum fluctuation values ​​are all 0, it means that the current data of each channel of the cabinet remains unchanged, and the abnormal situation may be caused by a hardware failure such as a virtual point or a collector being stuck.

[0118] At this time, the electronic equipment may generate an alarm message: Cabinet X may have a hardware failure such as a virtual point or a collector stuck. Please check!

[0119] In one embodiment, the execution order of the above steps 520 to 540 is not limited, for example, they can be executed in parallel or in sequence. Figure 5 By analyzing the magnitude of each data point in the current data separately, it is possible to identify whether the target current of the corresponding target cabinet has a power outage or hardware failure, thereby improving the efficiency of fault identification.

[0120] Step 550 : Calculate the correlation value between each current data and each adjacent current data of the current data to form a correlation sequence for each current data.

[0121] Specifically, according to the above formula:

[0122] To calculate the correlation value, a correlation sequence of each current data is obtained.

[0123] Step 560 : Select first target current data and second target current data from the plurality of current data.

[0124] Step 570: Determine whether the first target correlation value in the first correlation sequence is a larger value in the first correlation sequence; and determine whether the second target correlation value in the second correlation sequence is a larger value in the second correlation sequence.

[0125] The first correlation sequence is a correlation sequence of the first target current data; the second correlation sequence is a correlation sequence of the second target current data. The larger value may include a maximum value and a second largest value. The absolute value of the difference between the second largest value and the maximum value is less than a difference threshold.

[0126] Step 580: When both values ​​are larger, it is determined that the target cabinet power supply circuit is correctly wired.

[0127] Step 590: When either of the two values ​​is not a larger value, the object corresponding to the largest correlation value in the correlation sequence of each current data is taken as the most correlated object.

[0128] Specifically, the object corresponding to the largest correlation value in the correlation sequence may be regarded as the most relevant object.

[0129] Step 595: If the most relevant object of the two-way current data of the target cabinet is the same-way current data of the adjacent cabinet, and the most relevant object of the two-way current data of the adjacent cabinet is also the same-way current data of the target cabinet, then the wiring error type is determined to be that the target cabinet and its adjacent cabinet have different power supplies connected incorrectly; if the most relevant object of the two-way current data of the target cabinet is the different-way current data of the adjacent cabinet, and the most relevant object of the two-way current data of the adjacent cabinet is also the different-way current data of the target cabinet, then the wiring error type is determined to be that the target cabinet and its adjacent cabinet have the same power supply connected incorrectly.

[0130] It can be understood that the specific understanding and explanation of step 590 and step 595 can be found in the above steps 410 to 430.

[0131] In another embodiment, the following conditions 1 to 4 are detected to determine whether they are satisfied or established. If at least three of the following conditions 1 to 4 occur, it indicates that the power supply lines of cabinets X and Y are incorrectly connected. Alternatively, if any one of the following conditions 1 to 2 and any one of the following conditions 3 to 4 occur, it also indicates that the power supply lines of cabinets X and Y are incorrectly connected. Y can be X+1 or X-1. For example, if all of the following conditions 1 to 4 are satisfied, it indicates that the power supply lines of cabinets X and Y are incorrectly connected.

[0132] Case 1: M_c["X_A"]=="Y_A";

[0133] Case 2: Mc["X_B"] == "Y_B";

[0134] Case 3: M_c["Y_A"]=="X_A";

[0135] Case 4: M_c[“Y_B”] == “X_B”.

[0136] If the above conditions are met, it can be determined that cabinet X and its adjacent cabinet Y have different power connections incorrectly. For example, the power connector on cabinet X that should be connected to power supply A is connected to power supply B on cabinet Y, and / or the power connector on cabinet X that should be connected to power supply B is connected to power supply A on cabinet Y.

[0137] At this time, the electronic device may output a specific warning prompt: there is an error, X_A and Y_B are connected incorrectly or X_B and Y_A are connected incorrectly.

[0138] In another embodiment, the following conditions 5 to 8 are detected. If at least three of the following conditions 5 to 8 occur, it indicates that the power supply of cabinet X and cabinet Y is incorrectly connected. Alternatively, if any one of the following conditions 5 to 6 and any one of the following conditions 7 to 8 occur, it also indicates that the power supply of cabinet X and the adjacent cabinet Y is incorrectly connected. For example, if all of the following conditions 5 to 8 are met, it indicates that the power supply of cabinet X and cabinet Y is incorrectly connected.

[0139] Case 5: Mc["X_A"] == "Y_B";

[0140] Case 6: Mc["X_B"] == "Y_A";

[0141] Case 7: Mc["Y_A"] == "X_B";

[0142] Case 8: M_c[“Y_B”] == “X_A”.

[0143] If the above conditions are met, it can be determined that the power supply connections for cabinet X and its adjacent cabinet Y are incorrect. For example, the power connector on cabinet X that should be connected to cabinet X's power supply A is connected to cabinet Y's power supply A, and / or the power connector on cabinet X that should be connected to cabinet X's power supply B is connected to cabinet Y's power supply B.

[0144] Likewise, electronic devices can also output specific warning prompts.

[0145] The current-based cabinet routing anomaly detection method in this application can timely locate possible anomalies in specific cabinets by collecting, processing and analyzing a large amount of current data, such as collection failure, hardware failure, power outage, wiring error and other problems, which is of great significance to the stable operation of the data center.

[0146] In one embodiment, Figure 6 As shown, a current-based cabinet routing anomaly detection device is provided, the device comprising:

[0147] The current data acquisition module 610 is used to acquire the current data of each channel of each cabinet in a certain time period after the cabinet is powered on;

[0148] A correlation calculation module 620 is configured to calculate a correlation value between each current data item and each adjacent current data item of the current data item, thereby forming a correlation sequence for each current data item. Adjacent current data items represent, relative to one current data item of one cabinet, current data items of different current data items in the same cabinet and current data items of cabinets adjacent to the same cabinet.

[0149] The fault identification module 630 is configured to determine whether there is a wiring error in the power supply lines of the cabinets in the plurality of cabinets based on a correlation sequence of one or more current data.

[0150] In one embodiment, the fault identification module 630 is further configured to determine that the power line wiring of the cabinet is incorrect when the correlation value between two current data of the same cabinet is a smaller value in the correlation sequence of any current data of the same cabinet.

[0151] In one embodiment, the fault identification module 630 is also used to select first target current data and second target current data from multiple current data; based on the first correlation sequence of the first target current data and the second correlation sequence of the second target current data, determine whether there is a power line wiring error in the target cabinet, and the first target current data and the second target current data are two current data of the target cabinet respectively.

[0152] In one embodiment, the fault identification module 630 is also used to determine whether the first target correlation value in the first correlation sequence is a larger value in the first correlation sequence, and the first target correlation value is the correlation value of the second target current data relative to the first target current data; determine whether the second target correlation value in the second correlation sequence is a larger value in the second correlation sequence, and the second target correlation value is the correlation value of the first target current data relative to the second target current data; when both are larger values, it is determined that the target cabinet power line wiring is correct; when any one of the two is not a larger value, it is determined that the target cabinet power line wiring is incorrect.

[0153] In one embodiment, the fault identification module 630 is also used to take the object corresponding to the maximum correlation value in each correlation sequence as the most relevant object; if the most relevant object of the two-way current data of the target cabinet is the same-way current data of the adjacent cabinet, and the most relevant object of the two-way current data of the adjacent cabinet is also the same-way current data of the target cabinet, then it is determined that the different power supplies of the target cabinet and the adjacent cabinet are connected incorrectly; if the most relevant object of the two-way current data of the target cabinet is the different-way current data of the adjacent cabinet, and the most relevant object of the two-way current data of the adjacent cabinet is also the different-way current data of the target cabinet, then it is determined that the same-way power supplies of the target cabinet and the adjacent cabinet are connected incorrectly.

[0154] In one embodiment, the fault identification module 630 is further configured to identify whether there are persistent missing values ​​or 0 values ​​in each current data of each cabinet. If so, it is determined that the corresponding cabinet has a power outage or data collection failure.

[0155] In one embodiment, the fault identification module 630 is further used to identify whether there are multiple consecutive data points in each current data of each cabinet that are all smaller than a first threshold. If so, it is determined that the corresponding cabinet is not configured with a computing device or is in a test condition.

[0156] In one embodiment, the fault identification module 630 is also used to identify whether the maximum fluctuation value between data points in each current data of each cabinet is less than the second threshold and not 0 at the same time. If so, it is determined that the corresponding cabinet is not configured with computing equipment or is in a test condition.

[0157] In one embodiment, a computer-readable storage medium is provided, on which executable instructions are stored. When the instructions are executed by a processor, the processor executes the steps in the above-mentioned method embodiments.

[0158] In one embodiment, an electronic device is also provided, comprising one or more processors; a memory, wherein one or more programs are stored in the memory, wherein when the one or more programs are executed by one or more processors, the one or more processors execute the steps in the above-mentioned method embodiments.

[0159] In one embodiment, an electronic device is provided, which may be the device deployed with the intelligent building management system. Figure 7 As shown, electronic device 700 includes a central processing unit (CPU) 701, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 702 or a program loaded from a storage portion 708 into a random access memory (RAM) 703. Various programs and data required for the operation of electronic device 700 are also stored in RAM 703. CPU 701, ROM 702, and RAM 703 are connected to each other via a bus 704. An input / output (I / O) interface 705 is also connected to bus 704.

[0160] The following components are connected to the I / O interface 705: an input section 706 including a keyboard, a mouse, and the like; an output section 707 including devices such as a cathode ray tube (CRT), a liquid crystal display (LCD), and a speaker; a storage section 708 including a hard disk; and a communication section 709 including a network interface card such as a LAN card or a modem. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the I / O interface 705 as needed. A removable medium 711, such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, is installed in the drive 710 as needed, so that computer programs read therefrom can be installed into the storage section 708 as needed.

[0161] In particular, according to embodiments of the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present application include a computer program product comprising a computer-readable medium carrying instructions. In such embodiments, the instructions can be downloaded and installed from a network via a communication portion 709 and / or installed from a removable medium 711. When the instructions are executed by a central processing unit (CPU) 701, the various method steps described in the present invention are performed.

[0162] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some or all of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application.

[0163] In addition, those skilled in the art will understand that although some embodiments herein include certain features included in other embodiments but not other features, the combination of features of different embodiments is intended to be within the scope of the present application and form different embodiments. The information disclosed in this background technology section is intended only to deepen the understanding of the overall background technology of this application and should not be regarded as an admission or any form of implication that the information constitutes prior art already known to those skilled in the art.

Claims

1. A method for detecting cabinet routing anomalies based on current, characterized in that: The method comprises: After the cabinet is powered on, obtain the current data of each circuit in each cabinet in a certain time period; Calculating the correlation value between each current data and each adjacent current data of the current data to form a correlation sequence for each current data, wherein the adjacent current data represents, with respect to one current data of one cabinet, different current data in the same cabinet and current data of cabinets adjacent to the same cabinet; It is determined whether there is a wiring error of power lines of the cabinets in the plurality of cabinets based on a correlation sequence of the one or more current data.

2. The anomaly detection method according to claim 1, wherein: Each cabinet includes two channels of current data, and the adjacent current data of one channel of current data of any cabinet include the other channel of current data of the cabinet and the two channels of current data of the two cabinets adjacent to the cabinet. The correlation sequence of each current data includes the correlation value between the current data and each adjacent current data of the current data.

3. The method according to claim 2, characterized in that The determining whether there is a power line wiring error in the cabinets based on the correlation sequence of the one or more current data includes: When the correlation value between two current data of the same cabinet is a smaller value in the correlation sequence of any current data of the same cabinet, it is determined that the power line wiring of the cabinet is incorrect.

4. The anomaly detection method according to claim 2, wherein: The determining whether there is a power line wiring error in the cabinets based on the correlation sequence of the one or more current data includes: selecting first target current data and second target current data from a plurality of current data; Based on the first correlation sequence of the first target current data and the second correlation sequence of the second target current data, it is determined whether the target cabinet has a power line wiring error, and the first target current data and the second target current data are two current data of the target cabinet respectively.

5. The anomaly detection method according to claim 4, characterized in that: The determining whether a power line wiring error exists in the target cabinet based on a first correlation sequence of the first target current data and a second correlation sequence of the second target current data includes: determining whether a first target correlation value in the first correlation sequence is a larger value in the first correlation sequence, the first target correlation value being a correlation value of the second target current data relative to the first target current data; determining whether a second target correlation value in the second correlation sequence is a larger value in the second correlation sequence, the second target correlation value being a correlation value of the first target current data relative to the second target current data; When both values ​​are larger, it is determined that the target cabinet power line wiring is correct; When either value is not a larger value, it is determined that the target cabinet power supply circuit is incorrectly wired.

6. The anomaly detection method according to claim 1, wherein: Each cabinet includes two channels of current data; and determining whether there is a power line wiring error in the cabinets based on the correlation sequence includes: The object corresponding to the maximum correlation value in each correlation sequence is regarded as the most relevant object; If the most relevant object of the two-way current data of the target cabinet is the current data of the same way of the adjacent cabinet, and the most relevant object of the two-way current data of the adjacent cabinet is also the current data of the same way of the target cabinet, it is determined that the different power supplies of the target cabinet and the adjacent cabinet are connected incorrectly; If the most relevant object of the two-way current data of the target cabinet is the different-way current data of the adjacent cabinet, and the most relevant object of the two-way current data of the adjacent cabinet is also the different-way current data of the target cabinet, it is determined that the same-way power supply of the target cabinet and the adjacent cabinet is connected incorrectly.

7. The anomaly detection method according to claim 1, wherein: Calculate the current data I of the i-way current in cabinet m according to the following formula m_i The j-way current data I of cabinet n n_j The correlation value C m_i / n_j : K represents the number of data points in the current data, m_i k Indicates current data I m_i The value of the kth data point in n_j k Indicates current data I n_j The value of the kth data point in .

8. The method according to any one of claims 1 to 7, characterized in that After obtaining the current data of each channel of each cabinet in the plurality of cabinets within a certain time period, the method further includes: Identify whether there are persistent missing values ​​or zero values ​​in the current data of each channel of each cabinet. If so, determine that the corresponding cabinet has a power outage or data collection failure; and / or Identify whether there are multiple consecutive data points in each current data of each cabinet that are all smaller than a first threshold value. If so, determine that the corresponding cabinet is not configured with a computing device or is in a test condition; and / or Identify whether the maximum fluctuation value between data points in each current data of each cabinet is less than the second threshold and not 0 at the same time. If so, determine that the corresponding cabinet is not configured with computing equipment or is in a test condition.

9. A cabinet routing anomaly detection device based on current, characterized in that: The device comprises: A current data acquisition module is used to obtain the current data of each channel of each cabinet in a certain time period after the cabinet is powered on; a correlation calculation module, configured to calculate a correlation value between each current data and each adjacent current data of the current data, to form a correlation sequence for each current data, wherein the adjacent current data represents, relative to one current data of one cabinet, different current data in the same cabinet and current data of cabinets adjacent to the same cabinet; The fault identification module is configured to determine whether there is a wiring error in the power supply lines of the cabinets in the plurality of cabinets based on a correlation sequence of one or more current data.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores executable instructions, which, when executed by a processor, enable the processor to perform the method according to any one of claims 1 to 8.

11. An electronic device, characterized in that: include: one or more processors; A memory for storing one or more programs, which, when executed by the one or more processors, causes the one or more processors to perform the method according to any one of claims 1 to 8.

Images