WO2014132611A1 - システム分析装置、及び、システム分析方法 - Google Patents
システム分析装置、及び、システム分析方法 Download PDFInfo
- Publication number
- WO2014132611A1 WO2014132611A1 PCT/JP2014/000949 JP2014000949W WO2014132611A1 WO 2014132611 A1 WO2014132611 A1 WO 2014132611A1 JP 2014000949 W JP2014000949 W JP 2014000949W WO 2014132611 A1 WO2014132611 A1 WO 2014132611A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- correlation
- metric
- detection sensitivity
- destruction
- pair
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
- G06F11/079—Root cause analysis, i.e. error or fault diagnosis
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
- G05B23/0221—Preprocessing measurements, e.g. data collection rate adjustment; Standardization of measurements; Time series or signal analysis, e.g. frequency analysis or wavelets; Trustworthiness of measurements; Indexes therefor; Measurements using easily measured parameters to estimate parameters difficult to measure; Virtual sensor creation; De-noising; Sensor fusion; Unconventional preprocessing inherently present in specific fault detection methods like PCA-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
- G06F11/0706—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
- G06F11/0751—Error or fault detection not based on redundancy
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
- G06F11/0766—Error or fault reporting or storing
- G06F11/0772—Means for error signaling, e.g. using interrupts, exception flags, dedicated error registers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
- G06F11/3447—Performance evaluation by modeling
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/06—Management of faults, events, alarms or notifications
- H04L41/0631—Management of faults, events, alarms or notifications using root cause analysis; using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
Definitions
- the present invention relates to a system analysis apparatus and a system analysis method.
- Patent Document 1 describes an example of an operation management system that models a system using time series information of system performance and determines a factor such as a failure or abnormality of the system using the generated model. .
- the operation management system described in Patent Document 1 determines a correlation function of a system by determining a correlation function representing a correlation of each pair of a plurality of metrics based on measurement values of a plurality of metrics of the system. Generate.
- the operation management system detects the destruction of the correlation (correlation destruction) using the generated correlation model, and determines the failure factor of the system based on the correlation destruction.
- the technique for analyzing the state of the system based on correlation destruction is called invariant relation analysis.
- Patent Document 2 discloses a method for determining a failure point based on a correlation between points when a physical quantity at a plurality of points in the process changes from a reference point.
- FIG. 9 is a diagram showing a determination example of an abnormal factor in the invariant relation analysis of Patent Document 1.
- each node indicates a metric, and an arrow between the metrics indicates a correlation.
- the node indicated by the thick line indicates a metric in which an abnormality has occurred (an abnormality factor metric), and the thick arrow indicates a correlation in which correlation destruction is detected.
- a correlation breakdown is detected in one correlation (between metrics A and C) due to an abnormality in metric A.
- the metric of the abnormal factor is determined based on the ratio of the number of correlations in which correlation destruction is detected with respect to the number of all correlations (hereinafter referred to as the ratio of correlation destruction). The method is used.
- the ratio 1/2 of the correlation destruction related to the metric C is larger than the ratio 1/3 of the correlation destruction related to the metric A, and it is erroneously determined that the metric C is an abnormal factor.
- An object of the present invention is to solve the above-described problems and provide a system analysis apparatus and a system analysis method capable of accurately determining an abnormal factor in invariant relation analysis.
- a system analysis apparatus includes a correlation model storage unit that stores a correlation model indicating a correlation between a pair of metrics in a system, and a correlation in which a correlation breakdown is detected among the correlations included in the correlation model. Based on the detection sensitivity calculated for each metric related to the relationship, the detection sensitivity indicating the likelihood of occurrence of correlation destruction in each correlation related to the metric at the time of abnormality of the metric, to extract the metric of the candidate of the abnormal factor, An anomaly factor extracting means.
- the system analysis method stores a correlation model indicating a correlation between a pair of metrics in the system, and each metric related to the correlation in which the correlation destruction is detected among the correlations included in the correlation model. Based on the detection sensitivity calculated with respect to the detection sensitivity indicating the likelihood of occurrence of correlation destruction in each correlation related to the metric when the metric is abnormal, a metric of candidate abnormal factors is extracted.
- a computer stores a correlation model indicating a correlation of a pair of metrics in the system, and a correlation destruction among the correlations included in the correlation model is detected. Based on the detection sensitivity calculated for each metric related to the correlation and indicating the likelihood of correlation destruction occurring in each correlation related to the metric when the metric is abnormal, a metric for candidate abnormal factors is extracted. Stores the program that executes the process.
- the effect of the present invention is that the abnormal factor can be accurately determined in the invariant relation analysis.
- FIG. 1 It is a figure which shows the example of the correlation model 122 and the detection sensitivity in the 2nd Embodiment of this invention. It is a figure which shows the comparative example of the example of a detection of correlation destruction, and the detection sensitivity in the 2nd Embodiment of this invention. It is a figure which shows the example of determination of the abnormal factor in the invariant relationship analysis of patent document 1.
- FIG. 1 shows the example of the correlation model 122 and the detection sensitivity in the 2nd Embodiment of this invention. It is a figure which shows the comparative example of the example of a detection of correlation destruction, and the detection sensitivity in the 2nd Embodiment of this invention. It is a figure which shows the example of determination of the abnormal factor in the invariant relationship analysis of patent document 1.
- FIG. 2 is a block diagram showing a configuration of the system analysis apparatus 100 according to the first embodiment of the present invention.
- system analysis apparatus 100 is connected to a monitored system including one or more monitored apparatuses 200.
- the monitored device 200 is a device that constitutes an IT system, such as various server devices and network devices.
- the monitored device 200 measures the measurement data (measured values) of the performance values of the plurality of types of the monitored device 200 at regular intervals, and transmits them to the system analysis device 100.
- the performance value item for example, the usage rate and usage of computer resources and network resources such as CPU (Central Processing Unit) usage rate, memory usage rate, disk access frequency, etc. are used.
- CPU Central Processing Unit
- a combination of the monitored device 200 and the item of performance value is a metric (performance index), and a set of a plurality of metric values measured at the same time is performance information.
- Metrics are represented by integers and decimal numbers.
- a metric corresponds to an “element” that is a generation target of a correlation model in Patent Document 1.
- the system analysis apparatus 100 generates a correlation model 122 of the monitored apparatus 200 based on the performance information collected from the monitored apparatus 200, and analyzes the state of the monitored apparatus 200 using the generated correlation model 122. .
- the system analysis apparatus 100 includes a performance information collection unit 101, a correlation model generation unit 102, a correlation destruction detection unit 103, an abnormal factor extraction unit 104, a performance information storage unit 111, a correlation model storage unit 112, a correlation destruction storage unit 113, and A detection sensitivity storage unit 114 is included.
- the performance information collection unit 101 collects performance information from the monitored device 200.
- the performance information storage unit 111 stores the time series change of the performance information collected by the performance information collection unit 101 as performance series information.
- the correlation model generation unit 102 generates a correlation model 122 of the monitored system based on the performance sequence information.
- the correlation model 122 includes a correlation function (or prediction formula) representing the correlation between each pair of pairs of metrics.
- a correlation function is a function that predicts one value of a pair of metrics from both time series of the pair or the other time series.
- a target metric a metric predicted by the correlation function
- the other metric is referred to as a non-target metric.
- the correlation model generation unit 102 uses the system identification process for performance information in a predetermined modeling period to calculate the metric y (t), u (t)
- the correlation function f (y, u) is determined as in the equation.
- metrics y (t) and u (t) are an objective metric and a non-objective metric, respectively.
- An, bm, c, N, K, and M are determined so that the value of the prediction accuracy (fitness) of the correlation function expressed by Equation 2 is maximized.
- the correlation model generation unit 102 may use a set of correlation functions with a prediction accuracy equal to or higher than a predetermined value as the correlation model 122.
- FIG. 4 is a diagram showing an example of the correlation model 122 and detection sensitivity in the first embodiment of the present invention.
- the correlation model 122 is shown by a graph including nodes and arrows.
- each node indicates a metric
- an arrow between the metrics indicates a correlation.
- the metric at the end of the arrow corresponds to the target metric.
- one metric (hereinafter referred to as metrics A to D) exists in each of the monitored devices 200 having the device identifiers A to D, and the correlation is performed for each pair of the metrics A to D.
- a relationship is defined.
- one correlation function that predicts one metric of the pair is defined for the correlation of each pair of metrics.
- the correlation model storage unit 112 stores the correlation model 122 generated by the correlation model generation unit 102.
- Correlation destruction detector 103 detects the correlation destruction of the correlation included in correlation model 122 for the newly input performance information.
- the correlation destruction detection unit 103 detects the correlation destruction for each metric pair.
- the difference (prediction error) between the predicted value of the target metric obtained by inputting the measured value of the metric into the correlation function and the measured value of the target metric is equal to or greater than a predetermined threshold, Detect as correlation destruction of pair correlation.
- the correlation destruction storage unit 113 stores correlation destruction information indicating the correlation in which the correlation destruction is detected.
- FIG. 5 is a diagram showing a correlation destruction detection example and a detection sensitivity comparison example in the first embodiment of the present invention.
- thick arrows indicate correlations in which correlation destruction is detected in the correlation model 122 of FIG. 4.
- nodes indicated by bold lines indicate metrics in which an abnormality has occurred (an abnormality factor metric).
- the correlation destruction occurs in the correlation between the metric A and the metric C.
- the abnormality factor extraction unit 104 calculates the detection sensitivity of each correlation included in the correlation model 122.
- the detection sensitivity indicates the magnitude of the influence of the metric abnormality related to the correlation on the predicted value, that is, the likelihood of occurrence of correlation destruction in the correlation when the metric is abnormal.
- the correlation is expressed by a correlation function such as Equation 1 above, when a physical failure related to one of the metric pairs occurs, the prediction error of the predicted value of the target metric of the correlation function is: There is a tendency to increase in either positive or negative direction. In this case, the likelihood of occurrence of correlation destruction (detection sensitivity) in the correlation when the metric is abnormal can be approximately expressed by the sum of the coefficients of the correlation function representing the correlation.
- a value obtained by normalizing the sum of the coefficients of the correlation function with a prediction error threshold applied when determining correlation destruction is defined as detection sensitivity.
- the detection sensitivity is calculated as follows.
- the detection sensitivity Sy for the target metric y is calculated by dividing the sum of the coefficients by which the target metric y is multiplied in the correlation function f (y, u) by the prediction error threshold, as shown in Equation 3.
- the detection sensitivity Su for the non-objective metric u is calculated by dividing the sum of the coefficients by which the metric u is multiplied in the correlation function f (y, u) by the prediction error threshold, as shown in Equation 4.
- Threshold is a prediction error threshold applied when determining the correlation destruction using the correlation function f (y, u).
- the value of the threshold is determined by the correlation model generation unit 102 based on the maximum value of the prediction error or the standard deviation for the performance information in the modeling period.
- a threshold value may be set for each correlation function by an administrator or the like.
- the abnormal factor extraction unit 104 further extracts a candidate metric of an abnormal factor using the detection sensitivity of each correlation related to the metric calculated for each metric related to the correlation in which the correlation destruction is detected.
- the detection sensitivity storage unit 114 stores the detection sensitivity calculated by the abnormality factor extraction unit 104.
- the system analysis apparatus 100 may be a computer that includes a CPU and a storage medium that stores a program and that operates under control based on the program. Further, the performance information storage unit 111, the correlation model storage unit 112, the correlation destruction storage unit 113, and the detection sensitivity storage unit 114 may be configured as individual storage media or a single storage medium.
- FIG. 3 is a flowchart showing the operation of the system analysis apparatus 100 according to the first embodiment of the present invention.
- a correlation model 122 as shown in FIG. 4 is generated by the correlation model generation unit 102 and stored in the correlation model storage unit 112. Further, it is assumed that the detection sensitivity as shown in FIG. 4 is calculated by the abnormality factor extraction unit 104 and stored in the detection sensitivity storage unit 114.
- the correlation destruction detection unit 103 detects the correlation destruction of the correlation included in the correlation model 122 using the performance information newly collected by the performance information collection unit 101 (step S101).
- the correlation destruction detection unit 103 detects the correlation destruction with respect to newly collected performance information as shown in FIG.
- the abnormal factor extraction unit 104 selects one of the metrics included in the correlation model 122 (step S102).
- the abnormality factor extraction unit 104 selects one of the correlations related to the selected metric (Ste S104). Then, when the selected metric is the objective metric of the correlation function of the selected correlation (step S105 / Y), the abnormality factor extraction unit 104 detects the detection sensitivity for the objective metric of the correlation function from the detection sensitivity storage unit 114. To get. Further, when the selected metric is not the objective metric of the correlation function of the selected correlation (step S105 / N), the abnormality factor extraction unit 104 detects the detection sensitivity for the non-target metric of the correlation function from the detection sensitivity storage unit 114. To get. The abnormality factor extraction unit 104 repeats the processing from step S104 to S107 for all correlations related to the selected metric (step S108).
- the abnormality factor extraction unit 104 compares the detection sensitivities acquired for each correlation related to the selected metric, and determines whether or not correlation destruction is detected in the correlation having the highest detection sensitivity (step S109). . If correlation destruction is detected in the correlation having the highest detection sensitivity in step S109 (step S109 / Y), the abnormality factor extraction unit 104 determines that the selected metric is a candidate for the abnormality factor.
- the abnormality factor extraction unit 104 repeats the processing from step S102 to step S110 for all metrics included in the correlation model 122 (step S111).
- the abnormality factor extraction unit 104 outputs an identifier of the metric determined as a candidate for the abnormality factor to an administrator or the like by an output unit (not shown) (step S112).
- the abnormality factor extraction unit 104 outputs the metric A as an abnormality factor candidate.
- FIG. 1 is a block diagram showing a characteristic configuration of the first embodiment of the present invention.
- the system analysis apparatus 100 includes a correlation model storage unit 112 and an abnormal factor extraction unit 104.
- the correlation model storage unit 112 stores a correlation model 122 indicating the correlation between metric pairs in the system.
- the abnormal factor extraction unit 104 extracts candidate metrics of abnormal factors based on the detection sensitivity calculated for each metric related to the correlation in which the correlation destruction is detected among the correlations included in the correlation model 122. To do.
- the detection sensitivity indicates the likelihood of occurrence of correlation destruction in each correlation related to the metric when each metric is abnormal.
- the abnormal factor can be accurately determined in the invariant relation analysis.
- the reason is that the abnormal factor extraction unit 104 does not select all the metrics related to the correlation in which the correlation destruction is detected as candidates for abnormal factors, but further narrows down the metrics of candidate abnormal factors. That is, the abnormality factor extraction unit 104 narrows down the metric of candidate abnormal factors based on the detection sensitivity calculated for each metric related to the correlation in which the correlation destruction is detected.
- the detection sensitivity indicates the likelihood of occurrence of correlation destruction in each correlation related to the metric when each metric is abnormal.
- the detection sensitivity of the larger of the detection sensitivities of the two correlation functions is increased. This method is different from the first embodiment of the present invention in that the abnormal factor candidates are extracted.
- the configuration of the system analysis apparatus 100 in the second embodiment of the present invention is the same as that of the first embodiment (FIG. 2) of the present invention.
- FIG. 7 is a diagram illustrating an example of the correlation model 122 and the detection sensitivity in the second embodiment of the present invention.
- the correlation model 122 of FIG. 7 two correlation functions for predicting each metric of the pair are defined.
- the anomaly factor extraction unit 104 extracts the anomaly factor metric using the larger detection sensitivity of the two correlation functions representing each correlation.
- FIG. 6 is a flowchart showing the operation of the system analyzer 100 in the second embodiment of the present invention.
- the operation of the second embodiment of the present invention is the same as that of the first embodiment of the present invention except for the detection sensitivity acquisition process (steps S205 and S206 in FIG. 6) by the abnormality factor extraction unit 104. .
- a correlation model 122 as shown in FIG. 7 is generated by the correlation model generation unit 102 and stored in the correlation model storage unit 112. Further, it is assumed that the detection sensitivity as shown in FIG. 7 is calculated by the abnormality factor extraction unit 104 and stored in the detection sensitivity storage unit 114.
- FIG. 8 is a diagram showing a correlation destruction detection example and a detection sensitivity comparison example in the second embodiment of the present invention.
- a correlation breakdown is detected for each of the two correlation functions for the correlation of each pair of metrics.
- the correlation destruction detection unit 103 detects the correlation destruction with respect to the newly collected performance information as shown in FIG.
- the anomaly factor extraction unit 104 acquires the detection sensitivity for the target metric of the correlation function having the selected metric as the target metric out of the two correlation functions of the selected correlation. Furthermore, the abnormality factor extraction unit 104 acquires the detection sensitivity of the correlation function that uses the selected metric as the non-object metric for the non-object metric (step S205). Then, the abnormality factor extraction unit 104 selects a set of the detection sensitivity of the larger detection sensitivity and the detection status of the correlation destruction (step S206).
- the abnormality factor extraction unit 104 outputs the metric A as an abnormality factor candidate.
- the abnormal factor extraction unit 104 uses the larger detection sensitivity of the detection sensitivities of the two correlation functions representing each correlation to extract the metric of the abnormal factor candidate.
- the detection sensitivity of the correlation function is calculated by Equation 3 and Equation 4, but if a large value is obtained according to the coefficient multiplied by the metric, the detection sensitivity is changed to another method. You may decide by.
- the abnormality factor extraction unit 104 may determine the detection sensitivity using a detection sensitivity conversion table for coefficients. Further, the detection sensitivity may be determined by a method other than using a coefficient as long as it is possible to indicate the likelihood of occurrence of correlation destruction when the metric is abnormal.
- the metric when correlation destruction is detected in the correlation having the highest detection sensitivity, the metric is determined as a candidate for an abnormal factor. If the candidate can be extracted, the abnormality factor candidate may be determined by another method. For example, the abnormality factor extraction unit 104 may determine a candidate for an abnormality factor based on a score that increases in accordance with the number of detected correlation destructions by a correlation function having high detection sensitivity.
- the monitored system is an IT system that includes a server device, a network device, and the like as the monitored device 200.
- the monitored system may be another system as long as a correlation model of the monitored system is generated and the abnormality factor can be determined by correlation destruction.
- the monitored system may be a plant system, a structure, a transportation device, or the like.
- the system analysis apparatus 100 generates the correlation model 122 using the values of various sensors as metrics, detects correlation destruction, and extracts abnormal factor candidates.
- the present invention can be applied to invariant relation analysis in which a cause of a system abnormality or failure is determined based on correlation destruction detected on a correlation model.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Computer Hardware Design (AREA)
- Environmental & Geological Engineering (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Automation & Control Theory (AREA)
- Debugging And Monitoring (AREA)
Abstract
Description
本発明の第1の実施の形態について説明する。
次に、本発明の第2の実施の形態について説明する。
101 性能情報収集部
102 相関モデル生成部
103 相関破壊検出部
104 異常要因抽出部
111 性能情報記憶部
112 相関モデル記憶部
113 相関破壊記憶部
114 検出感度記憶部
122 相関モデル
200 被監視装置
Claims (15)
- システムにおけるメトリックのペアの相関関係を示す相関モデルを記憶する相関モデル記憶手段と、
前記相関モデルに含まれる相関関係の内の相関破壊が検出された相関関係に係る各メトリックについて算出された、当該メトリックの異常時の当該メトリックに係る各相関関係における相関破壊の発生しやすさを示す検出感度をもとに、異常要因の候補のメトリックを抽出する、異常要因抽出手段と、
を備えるシステム分析装置。 - 前記異常要因抽出手段は、前記相関破壊が検出された相関関係に係る各メトリックについて、当該メトリックに係る相関関係の内の最も検出感度が高い相関関係に相関破壊が検出されている場合、当該メトリックを前記異常要因の候補と判定する、
請求項1に記載のシステム分析装置。 - 前記メトリックのペアの相関関係は、当該ペアの一方のメトリックの値を当該ペアの両方の時系列、または、当該ペアの他方の時系列から予測する相関関数で表され、
前記相関関係に係るメトリックに対する当該相関関係の検出感度は、当該相関関係の相関関数において当該メトリックに乗じる係数に応じて大きくなるように決定される、
請求項1または2に記載のシステム分析装置。 - 前記相関関係に係るメトリックに対する当該相関関係の検出感度は、さらに、当該相関関係の相関関数を用いて相関破壊を判定するときに適用される、予測誤差の閾値に応じて小さくなるように決定される、
請求項3に記載のシステム分析装置。 - 前記メトリックのペアの相関関係は、当該ペアの各々を予測する2つの相関関数により表され、
前記異常要因抽出手段は、前記相関破壊が検出された相関関係に係るメトリックの各相関関係を表す2つの相関関数の検出感度の内、大きい方の検出感度を用いて、前記異常要因の候補のメトリックを抽出する、
請求項3または4に記載のシステム分析装置。 - システムにおけるメトリックのペアの相関関係を示す相関モデルを記憶し、
前記相関モデルに含まれる相関関係の内の相関破壊が検出された相関関係に係る各メトリックについて算出された、当該メトリックの異常時の当該メトリックに係る各相関関係における相関破壊の発生しやすさを示す検出感度をもとに、異常要因の候補のメトリックを抽出する、
システム分析方法。 - 前記異常要因の候補のメトリックの抽出において、前記相関破壊が検出された相関関係に係る各メトリックについて、当該メトリックに係る相関関係の内の最も検出感度が高い相関関係に相関破壊が検出されている場合、当該メトリックを前記異常要因の候補と判定する、
請求項6に記載のシステム分析方法。 - 前記メトリックのペアの相関関係は、当該ペアの一方のメトリックの値を当該ペアの両方の時系列、または、当該ペアの他方の時系列から予測する相関関数で表され、
前記相関関係に係るメトリックに対する当該相関関係の検出感度は、当該相関関係の相関関数において当該メトリックに乗じる係数に応じて大きくなるように決定される、
請求項6または7に記載のシステム分析方法。 - 前記相関関係に係るメトリックに対する当該相関関係の検出感度は、さらに、当該相関関係の相関関数を用いて相関破壊を判定するときに適用される、予測誤差の閾値に応じて小さくなるように決定される、
請求項8に記載のシステム分析方法。 - 前記メトリックのペアの相関関係は、当該ペアの各々を予測する2つの相関関数により表され、
前記異常要因の候補のメトリックの抽出において、前記相関破壊が検出された相関関係に係るメトリックの各相関関係を表す2つの相関関数の検出感度の内、大きい方の検出感度を用いて、前記異常要因の候補のメトリックを抽出する、
請求項8または9に記載のシステム分析方法。 - コンピュータに、
システムにおけるメトリックのペアの相関関係を示す相関モデルを記憶し、
前記相関モデルに含まれる相関関係の内の相関破壊が検出された相関関係に係る各メトリックについて算出された、当該メトリックの異常時の当該メトリックに係る各相関関係における相関破壊の発生しやすさを示す検出感度をもとに、異常要因の候補のメトリックを抽出する、
処理を実行させるプログラムを格納する、コンピュータが読み取り可能な記録媒体。 - 前記異常要因の候補のメトリックの抽出において、前記相関破壊が検出された相関関係に係る各メトリックについて、当該メトリックに係る相関関係の内の最も検出感度が高い相関関係に相関破壊が検出されている場合、当該メトリックを前記異常要因の候補と判定する、
処理を実行させる請求項11に記載のプログラムを格納する、コンピュータが読み取り可能な記録媒体。 - 前記メトリックのペアの相関関係は、当該ペアの一方のメトリックの値を当該ペアの両方の時系列、または、当該ペアの他方の時系列から予測する相関関数で表され、
前記相関関係に係るメトリックに対する当該相関関係の検出感度は、当該相関関係の相関関数において当該メトリックに乗じる係数に応じて大きくなるように決定される、
請求項11または12に記載のプログラムを格納する、コンピュータが読み取り可能な記録媒体。 - 前記相関関係に係るメトリックに対する当該相関関係の検出感度は、さらに、当該相関関係の相関関数を用いて相関破壊を判定するときに適用される、予測誤差の閾値に応じて小さくなるように決定される、
請求項13に記載のプログラムを格納する、コンピュータが読み取り可能な記録媒体。 - 前記メトリックのペアの相関関係は、当該ペアの各々を予測する2つの相関関数により表され、
前記異常要因の候補のメトリックの抽出において、前記相関破壊が検出された相関関係に係るメトリックの各相関関係を表す2つの相関関数の検出感度の内、大きい方の検出感度を用いて、前記異常要因の候補のメトリックを抽出する、
処理を実行させる請求項13または14に記載のプログラムを格納する、コンピュータが読み取り可能な記録媒体。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2015502761A JP6183449B2 (ja) | 2013-02-26 | 2014-02-24 | システム分析装置、及び、システム分析方法 |
US14/766,880 US20150378806A1 (en) | 2013-02-26 | 2014-02-24 | System analysis device and system analysis method |
EP14756415.7A EP2963552B1 (en) | 2013-02-26 | 2014-02-24 | System analysis device and system analysis method |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2013035784 | 2013-02-26 | ||
JP2013-035784 | 2013-02-26 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2014132611A1 true WO2014132611A1 (ja) | 2014-09-04 |
Family
ID=51427890
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2014/000949 WO2014132611A1 (ja) | 2013-02-26 | 2014-02-24 | システム分析装置、及び、システム分析方法 |
Country Status (4)
Country | Link |
---|---|
US (1) | US20150378806A1 (ja) |
EP (1) | EP2963552B1 (ja) |
JP (1) | JP6183449B2 (ja) |
WO (1) | WO2014132611A1 (ja) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2018530803A (ja) * | 2015-07-14 | 2018-10-18 | サイオス テクノロジー コーポレーションSios Technology Corporation | コンピュータ環境における根本原因分析および修復のために機械学習原理を活用する装置および方法 |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103026344B (zh) * | 2010-06-07 | 2015-09-09 | 日本电气株式会社 | 故障检测设备、故障检测方法和程序记录介质 |
CN104137078B (zh) * | 2012-01-23 | 2017-03-22 | 日本电气株式会社 | 操作管理设备、操作管理方法和程序 |
JP6445859B2 (ja) * | 2014-12-16 | 2018-12-26 | 株式会社東芝 | プラント監視装置 |
US10581665B2 (en) * | 2016-11-04 | 2020-03-03 | Nec Corporation | Content-aware anomaly detection and diagnosis |
CN110225540A (zh) * | 2019-01-30 | 2019-09-10 | 北京中科晶上科技股份有限公司 | 一种面向集中式接入网的故障检测方法 |
WO2022018467A1 (en) * | 2020-07-22 | 2022-01-27 | Citrix Systems, Inc. | Determining changes in a performance of a server |
WO2023215903A1 (en) * | 2022-05-06 | 2023-11-09 | Mapped Inc. | Automatic link prediction for devices in commercial and industrial environments |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS6351936A (ja) | 1986-08-22 | 1988-03-05 | Hisayoshi Matsuyama | プロセスの異常診断方法 |
JP2006135412A (ja) * | 2004-11-02 | 2006-05-25 | Tokyo Gas Co Ltd | 遠隔監視システム |
JP4872944B2 (ja) | 2008-02-25 | 2012-02-08 | 日本電気株式会社 | 運用管理装置、運用管理システム、情報処理方法、及び運用管理プログラム |
JP2012242159A (ja) * | 2011-05-17 | 2012-12-10 | Internatl Business Mach Corp <Ibm> | システムの高い可用性のためにセンサデータを補間する方法、コンピュータプログラム、システム。 |
WO2013111560A1 (ja) * | 2012-01-23 | 2013-08-01 | 日本電気株式会社 | 運用管理装置、運用管理方法、及びプログラム |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6643613B2 (en) * | 2001-07-03 | 2003-11-04 | Altaworks Corporation | System and method for monitoring performance metrics |
US7107187B1 (en) * | 2003-11-12 | 2006-09-12 | Sprint Communications Company L.P. | Method for modeling system performance |
US8463899B2 (en) * | 2005-07-29 | 2013-06-11 | Bmc Software, Inc. | System, method and computer program product for optimized root cause analysis |
JP4872945B2 (ja) * | 2008-02-25 | 2012-02-08 | 日本電気株式会社 | 運用管理装置、運用管理システム、情報処理方法、及び運用管理プログラム |
US9195563B2 (en) * | 2011-03-30 | 2015-11-24 | Bmc Software, Inc. | Use of metrics selected based on lag correlation to provide leading indicators of service performance degradation |
US9298525B2 (en) * | 2012-12-04 | 2016-03-29 | Accenture Global Services Limited | Adaptive fault diagnosis |
-
2014
- 2014-02-24 JP JP2015502761A patent/JP6183449B2/ja active Active
- 2014-02-24 WO PCT/JP2014/000949 patent/WO2014132611A1/ja active Application Filing
- 2014-02-24 EP EP14756415.7A patent/EP2963552B1/en active Active
- 2014-02-24 US US14/766,880 patent/US20150378806A1/en not_active Abandoned
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS6351936A (ja) | 1986-08-22 | 1988-03-05 | Hisayoshi Matsuyama | プロセスの異常診断方法 |
JP2006135412A (ja) * | 2004-11-02 | 2006-05-25 | Tokyo Gas Co Ltd | 遠隔監視システム |
JP4872944B2 (ja) | 2008-02-25 | 2012-02-08 | 日本電気株式会社 | 運用管理装置、運用管理システム、情報処理方法、及び運用管理プログラム |
JP2012242159A (ja) * | 2011-05-17 | 2012-12-10 | Internatl Business Mach Corp <Ibm> | システムの高い可用性のためにセンサデータを補間する方法、コンピュータプログラム、システム。 |
WO2013111560A1 (ja) * | 2012-01-23 | 2013-08-01 | 日本電気株式会社 | 運用管理装置、運用管理方法、及びプログラム |
Non-Patent Citations (2)
Title |
---|
See also references of EP2963552A4 |
YUJI IZUMI ET AL.: "A Network-Status Evaluation Method Using Occurence Probability Matrices of Correlation Coefficients", THE TRANSACTIONS OF THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS, vol. J90-B, no. 7, 1 July 2007 (2007-07-01), pages 660 - 669 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2018530803A (ja) * | 2015-07-14 | 2018-10-18 | サイオス テクノロジー コーポレーションSios Technology Corporation | コンピュータ環境における根本原因分析および修復のために機械学習原理を活用する装置および方法 |
Also Published As
Publication number | Publication date |
---|---|
EP2963552A1 (en) | 2016-01-06 |
US20150378806A1 (en) | 2015-12-31 |
JPWO2014132611A1 (ja) | 2017-02-02 |
EP2963552A4 (en) | 2016-07-27 |
EP2963552B1 (en) | 2021-03-24 |
JP6183449B2 (ja) | 2017-08-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6183450B2 (ja) | システム分析装置、及び、システム分析方法 | |
JP6183449B2 (ja) | システム分析装置、及び、システム分析方法 | |
JP6394726B2 (ja) | 運用管理装置、運用管理方法、及びプログラム | |
US9658916B2 (en) | System analysis device, system analysis method and system analysis program | |
JP5874936B2 (ja) | 運用管理装置、運用管理方法、及びプログラム | |
JP5267748B2 (ja) | 運用管理システム、運用管理方法、及びプログラム | |
JP6196196B2 (ja) | ログ間因果推定装置、システム異常検知装置、ログ分析システム、及びログ分析方法 | |
JP5971395B2 (ja) | システム分析装置、及び、システム分析方法 | |
US10157113B2 (en) | Information processing device, analysis method, and recording medium | |
WO2015182072A1 (ja) | 因果構造推定システム、因果構造推定方法およびプログラム記録媒体 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 14756415 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14766880 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2014756415 Country of ref document: EP |
|
ENP | Entry into the national phase |
Ref document number: 2015502761 Country of ref document: JP Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |