WO2013011569A1 - ストリームデータ処理サーバ及びストリームデータ処理プログラムを記録した記録媒体 - Google Patents
ストリームデータ処理サーバ及びストリームデータ処理プログラムを記録した記録媒体 Download PDFInfo
- Publication number
- WO2013011569A1 WO2013011569A1 PCT/JP2011/066408 JP2011066408W WO2013011569A1 WO 2013011569 A1 WO2013011569 A1 WO 2013011569A1 JP 2011066408 W JP2011066408 W JP 2011066408W WO 2013011569 A1 WO2013011569 A1 WO 2013011569A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- data
- statistical value
- input
- time
- stream data
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/284—Relational databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
-
- 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
- H04L41/147—Network analysis or design for predicting network behaviour
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/60—Network structure or processes for video distribution between server and client or between remote clients; Control signalling between clients, server and network components; Transmission of management data between server and client, e.g. sending from server to client commands for recording incoming content stream; Communication details between server and client
- H04N21/63—Control signaling related to video distribution between client, server and network components; Network processes for video distribution between server and clients or between remote clients, e.g. transmitting basic layer and enhancement layers over different transmission paths, setting up a peer-to-peer communication via Internet between remote STB's; Communication protocols; Addressing
- H04N21/647—Control signaling between network components and server or clients; Network processes for video distribution between server and clients, e.g. controlling the quality of the video stream, by dropping packets, protecting content from unauthorised alteration within the network, monitoring of network load, bridging between two different networks, e.g. between IP and wireless
- H04N21/64723—Monitoring of network processes or resources, e.g. monitoring of network load
- H04N21/6473—Monitoring network processes errors
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/80—Generation or processing of content or additional data by content creator independently of the distribution process; Content per se
- H04N21/83—Generation or processing of protective or descriptive data associated with content; Content structuring
- H04N21/845—Structuring of content, e.g. decomposing content into time segments
- H04N21/8456—Structuring of content, e.g. decomposing content into time segments by decomposing the content in the time domain, e.g. in time segments
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/80—Generation or processing of content or additional data by content creator independently of the distribution process; Content per se
- H04N21/85—Assembly of content; Generation of multimedia applications
- H04N21/854—Content authoring
- H04N21/8547—Content authoring involving timestamps for synchronizing content
Definitions
- the present invention relates to stream data analysis processing using an approximate expression.
- the stream data processing system targets stream data.
- the stream data is a data sequence in chronological order that arrives without interruption.
- the query is a scenario indicating the data to be processed and the processing content, and is described using CQL (Continuous Query Language).
- sliding windows There are two types of sliding windows. Specifically, a number window that holds n time-series data immediately before the time to be processed and a time window that holds time-series data for n hours immediately before the time to be processed. It is a kind.
- the stream data processing system can analyze the state at the current time and cope with prediction of future data changes.
- a computer that processes stream data cuts out time-series data using a sliding window, and analyzes the relationship between time and target values (metrics) for the cut-out time-series data. To do.
- the computer calculates a relational expression (approximate expression) between time and metrics as an analysis result. As a result, a future change in value can be predicted.
- Equations (1) and (2) are solutions to the equation shown in Equation (3).
- An object of the present invention is to provide a stream data processing system that can calculate an approximate expression at low calculation cost without causing overflow even when the time value is large.
- a typical example of the invention disclosed in the present application is as follows. That is, a processor, a memory connected to the processor, a storage medium connected to the processor, and an interface connected to the processor and connected to another device are continued from the monitored computer system.
- a stream data processing server for processing incoming stream data wherein the stream data includes a plurality of data with time stamps, and the stream data processing server performs sliding according to a query registered in advance.
- a stream data processing unit that cuts out the data included in a target processing range from the stream data using a window and executes an analysis process on the cut-out data; and the stream data processing unit Using the collected data.
- An approximate expression calculation unit for calculating an approximate expression indicating a correspondence relationship between the timestamp and the value of the data, and calculating a predicted value of the data using the calculated approximate expression, and based on the calculated predicted value
- An abnormality detection unit that predicts the occurrence of an abnormality in the computer system, and the approximate expression calculation unit determines any time stamp of the extracted data as a time origin, and The time stamp is corrected to a relative time value from the origin of the determined time, and the approximate expression is calculated using the corrected time stamp and the data value.
- FIG. 1 is a block diagram illustrating a configuration example of a stream data processing system according to an embodiment of the present invention.
- the stream data processing system includes a stream data processing server 100, a monitoring target system, and a plurality of host computers 130 and 140.
- the stream data processing server 100 is connected to the monitoring target system 120 and the plurality of host computers 130 and 140 via the network 150.
- the network 150 may be a WAN, a LAN, or the like, but the present invention is not limited to the network connection type.
- the stream data processing server 100 receives the stream data transmitted from the monitoring target system 120 and processes the stream data according to the instructed query.
- the stream data includes a plurality of data 122 in time series order.
- the stream data processing server 100 includes a processor 101, a memory 102, a network interface 104, and a storage 105, and each component is connected via a bus 103.
- the processor 101 executes various processes by executing a program stored in the memory 102.
- the memory 102 stores a program executed by the processor 101 and information necessary for executing the program. Specifically, the memory 102 includes a stream data processing unit 110.
- the stream data processing unit 110 processes stream data.
- the stream data processing unit 110 reads a query group definition stored in the storage 105 at the start of processing, and configures a query graph based on the read query group definition.
- the stream data processing unit 110 executes processing according to the query graph.
- the stream data processing unit 110 includes an approximate expression calculation unit 112 and an abnormality detection unit 113.
- the approximate expression calculation unit 112 performs an analysis process on the plurality of data 122 cut out by the sliding window, and calculates an approximate expression.
- the data cut out by the sliding window is also referred to as target data.
- the sliding window is referred to as a window.
- the abnormality detection unit 113 detects the occurrence of an abnormality in the monitoring target system 120 using the calculated approximate expression and predicts the occurrence of the abnormality. For example, the abnormality detection unit 113 calculates a predicted value of metrics using an approximate expression, and determines whether the predicted value is equal to or greater than a predetermined threshold value.
- the network interface 104 is an interface for connecting to the network 150.
- the storage 105 stores stream data (data 122), a query 132, and other information.
- the storage 105 may be a storage medium such as an HDD or an SSD.
- the present invention is not limited to the type of storage medium.
- the monitoring target system 120 is a computer system composed of a plurality of computers (not shown).
- a system monitor 121 for monitoring data (metrics) to be monitored is executed on a computer (not shown) constituting the monitoring target system 120.
- the system monitor 121 collects necessary data from a computer (not shown) constituting the monitoring target system 120, and generates data 122 from the collected data. In addition, the system monitor 121 transmits the generated data 122 to the stream data processing server 100.
- the host computers 130 and 140 are computers used by users who use the stream data processing server 100, and include a processor (not shown), a memory (not shown), and a network interface (not shown).
- the abnormality monitoring query operation interface 131 is an interface for registering the abnormality monitoring query 132 and instructing the execution of the abnormality monitoring query 132.
- the stream data processing server 100 analyzes the abnormality monitoring query 132 and configures a query graph for executing the stream data processing.
- the stream data processing server 100 processes stream data according to the query graph.
- the abnormality monitoring process 141 is a process for displaying a processing result to the user and notifying an error or the like based on the result 142 transmitted from the stream data processing server 100.
- one computer may provide the abnormality monitoring query operation interface 131 and execute the abnormality monitoring process 141.
- 2 and 3 are explanatory views showing an example of an approximation method using a conventional least square method.
- an approximation method of the linear expression will be described.
- the horizontal axis x represents time
- the vertical axis y represents metrics. Metrics may be, for example, the usage rate of a processor to be monitored and the usage rate of a network band.
- the time means a time stamp given to the data.
- the coefficients a and b can be calculated using the equations (1) and (2).
- the values of the equations (4) to (7) may be obtained in order to calculate the coefficients a and b.
- S (t) the values shown in the equations (4) to (7) are not distinguished, they are described as a value S (t).
- the time origin is moved in accordance with the movement of the window. Specifically, the approximate expression calculation unit 112 first determines the time stamp of any one of the extracted data 122 as the origin. Next, the approximate expression calculation unit 112 corrects the time stamp of the cut out data 122 to the relative time from the corrected origin. Along with the time stamp correction processing described above, equations (4) to (7) are also changed.
- Equation (4) to (7) is deformed as shown Equation (8) to Formula (11). That is, the first problem can be solved by using the relative time. Further, when Expressions (8) to (11) are used, Expressions (1) and (2) can be expressed as Expressions (12) and (13).
- 4A and 4B are explanatory diagrams illustrating an example of an approximation method using the least square method according to the embodiment of the present invention.
- the approximate expression calculation unit 112 uses the time stamp of the latest data 411 in the window 410 as the origin.
- the origin O m of the time before moving the window 410 is the time stamp X t ⁇ 1 of the data 413.
- the time origin O m is moved to the time stamp X t of the data 423.
- the data 424 is deleted as the window moves.
- origin determination method of the present invention is not limited to the above-described method, and any time stamp included in the window 410 may be used as the origin.
- Equations (8) to (11) are transformed as shown in Equations (14) to (17).
- d t and z t are values representing a time difference, and are defined as in Expression (18) and Expression (19).
- the value S (t) is treated as a state value.
- each module of the approximate expression calculation unit 112 is configured to generate a time difference between the data input process and the data output process.
- FIG. 5 is an explanatory diagram illustrating a configuration of the approximate expression calculation unit 112 according to the embodiment of this invention.
- the approximate expression calculation unit 112 includes a data input unit 501, an initial state generation unit 502, a state value storage unit 503, a window data storage unit 504, a data storage unit 505, a state value update unit 506, a coefficient calculation unit 507, and a data output unit. 508.
- the data input unit 501 receives input of target data.
- the target data is input in the form of a time stamp and metrics. That is, target data in the format (x, y) is input.
- target data is input to the data input unit 501 one by one in chronological order.
- the initial state generation unit 502 When the target data is input to the approximate expression calculation unit 112 for the first time, the initial state generation unit 502 generates an initial value S (0) of the state value and outputs the generated initial value S (0). Specifically, the initial values of the equations (8) to (11) are “0”.
- the state value storage unit 503 stores the state value S (t). When target data is input for the first time, the state value storage unit 503 stores an initial value S (0).
- the window data storage unit 504 stores the target data cut out by the window. Further, the window data storage unit 504 executes data update processing as the window moves. Specifically, the following processing is executed.
- the window data storage unit 504 determines whether there is data to be deleted from the window as the window moves. That is, it is determined whether there is target data that is no longer included in the window as the window moves.
- the window data storage unit 504 When there is data to be deleted from the window, the window data storage unit 504 outputs the data to the state value update unit 506.
- data that is erased as the window moves is also referred to as erasure data.
- the processing of the window data storage unit 504 is different between the number window and the time window. Specific processing will be described later with reference to FIGS.
- the data storage unit 505 stores target data used for calculating the state value S (t).
- the data storage unit 505 stores target data that is one time before the target data received by the data input unit 501. For example, when the data input unit 501 receives target data (x 6 , y 6 ), the data storage unit 505 stores the target data (x 5 , y 5 ).
- State value update unit 506 calculates a state value S (t) using equations (14) to (17) when a value is input from each component unit.
- the coefficient calculation unit 507 substitutes the state value S (t) into the equations (12) and (13), and the coefficient a And b are calculated.
- the data output unit 508 generates an approximate expression based on the calculated coefficients a and b, and outputs the generated approximate expression to the abnormality detection unit 113.
- the state value storage unit 503, the window data storage unit 504, and the data storage unit 505 have a function of storing data in the storage area of the memory 102.
- the data input unit 501 When the target data is input, the data input unit 501 outputs the target data to the initial state generation unit 502, the window data storage unit 504, the data storage unit 505, and the state value update unit 506.
- the data input unit 501 generates a minute time delay and outputs the data to the data storage unit 505.
- the initial state generation unit 502 When the target data is input for the first time, the initial state generation unit 502 generates an initial state value S (0), and outputs the generated initial state value S (0) to the state value storage unit 503.
- the data storage unit 505 outputs the currently stored target data to the state value update unit 506, and stores new target data after a lapse of a minute time. Thus, the previous time-series target data can be held when processing the input target data.
- the state value storage unit 503 outputs the currently stored state value S (t) to the state value update unit 506.
- the state value update unit 506 substitutes the values input from the data input unit 501, the state value storage unit 503, the window data storage unit 504, and the data storage unit 505 into equations (14) to (17), and the window moves.
- the state value S (t + 1) after the calculation is calculated. Further, the state value update unit 506 outputs the calculated state value S (t + 1) to the state value storage unit 503 and the coefficient calculation unit 507.
- the state value update unit 506 generates a minute time delay and outputs the update value S (t + 1) to the state value storage unit 503. That is. After the minute time has elapsed, the state value S (t + 1) in the state value storage unit 503 is updated.
- the consistency between the input and the output is maintained by generating a minute time delay.
- recursive processing can be realized by providing a minute time delay when data is input to the state value storage unit 503 and the data storage unit 505.
- the minute time may be shorter than the time accuracy of the time stamp. For example, when the time accuracy of the time stamp is 1 millisecond, a delay of about 1 microsecond or 1 nanosecond may be generated.
- the input timing of the target data is the same as the generation timing of the disappearance data.
- the number of target data cut out by the number window is equal to or less than the number of data set in the number window, no disappearance data is generated.
- the state value update unit 506 it is necessary for the state value update unit 506 to change the calculation formula separately when the disappearance data is input and when the disappearance data is not input.
- FIG. 6 is a flowchart illustrating processing executed by the state value update unit 506 in the embodiment of the present invention.
- the state value update unit 506 calculates a time difference dt when target data is input from the data input unit 501 (step S601).
- the state value update unit 506 determines whether there is annihilation data (step S602).
- the state value update unit 506 determines whether extinction data is input from the window data storage unit 504. When extinction data is input, it is determined that there is extinction data.
- the state value update unit 506 calculates the time difference z t (step S603).
- the state value update unit 506 sets the time difference z t to “0” (step S605).
- the state value update unit 506 substitutes the values input from each unit into the equations (14) to (17), calculates each state value (step S604), and ends the process.
- FIG. 7 is an explanatory diagram showing data update timing of each component of the approximate expression calculation unit 112 according to the embodiment of the present invention.
- FIG. 7 illustrates an example of a number window in which seven pieces of data are cut out.
- the data input unit 501 receives input of target data (x 6 , y 6 ).
- the data storage unit 505 stores the target data (x 5 , y 5 ) in the previous time series.
- Data storage unit 505 outputs the target data (x 5, y 5) to the state value update section 506, the target data (x 5 from the target data (x 6, y 6) is inputted after a lapse of short time, y 5 ) to the target data (x 6 , y 6 ).
- the state value update unit 506 uses the target data (x 5 , y 5 ), the state value S (5), and the target data (x 6 , y 6 ).
- the state value S (6) is calculated. Further, the state value update unit 506 outputs the state value S (6) to the state value storage unit 503.
- the state value storage unit 503 stores the state value S (5) in the previous time series when the target data (x 6 , y 6 ) is input.
- the state value storage unit 503 updates the state value S (6) after a minute time has elapsed.
- the update process is executed in the same way when other target data is input.
- the window data storage unit 504 outputs the disappearance data to the state value update unit 506.
- the state value update unit 506 calculates the state value S (t) using a mathematical formula having a time difference z t of “0” until data (x 8 , y 8 ) is input.
- the state value update unit 506 needs to change the calculation formula used for the update process depending on the type of input data.
- FIG. 8 is a flowchart illustrating processing executed by the state value update unit 506 in the embodiment of the present invention.
- the state value update unit 506 determines whether or not the input data is extinction data (step S801). Since the determination process is the same as that in step S602, description thereof is omitted.
- the state value update unit 506 calculates the time difference z t (step S802), and further sets the time difference d t to “0” (step S803). ).
- the state value update unit 506 substitutes each value into the equations (14) to (17) to calculate the state value S (t) (step S804), and ends the process.
- the state value update unit 506 calculates the time difference dt (step S805), and further, the time difference z t is set to “0” (step S806).
- the state value update unit 506 substitutes each value into the equations (14) to (17) to calculate the state value S (t) (step S803), and ends the process.
- FIG. 9 is an explanatory diagram showing the data update timing of each component of the approximate expression calculation unit 112 according to the embodiment of the present invention.
- the input timing of the target data and the input timing of the disappearance data are different.
- the update method of the data input unit 501, the state value storage unit 503, and the data storage unit 505 is the same as that in the case of the number window, the description is omitted.
- the state value update unit 506 calculates the state value S (6) by executing the processes of step S805, step S806, and step S804.
- the state value update unit 506 calculates the state value S (7) by executing the processes of steps S802, S803, and S804.
- the present invention is not limited to a first-order approximation formula, but can be applied to a case of approximating a higher-order polynomial expression.
- the m-th order approximate expression may be obtained by solving a simultaneous equation having a coefficient that is a sum of powers of x and a sum of products of powers of x and y.
- representing the state value Sxm (t) is the m-th power sum of the time x at time x t as in Equation (21).
- the state value Sxm (t ⁇ 1) which is the sum of x to the m-th power at time x t ⁇ 1 , is expressed as in Equation (22).
- the sum of the product of the power of x and y can be calculated as an approximate expression using the expressions shown in Expression (24) and Expression (25). That is, the approximate expression calculation unit 112 can calculate the coefficient of the m-th order polynomial using a recursive and incremental calculation method with the configuration shown in FIG.
- the approximate expression calculation unit 112 can calculate the approximate expression while reducing the calculation cost.
Abstract
Description
以下、式(4)~式(7)に示す値を区別しない場合には値S(t)と記載する。
また、式(8)~式(11)を用いると式(1)及び式(2)は、式(12)及び式(13)のように表すことができる。
ここで、dt及びztは時刻差を表す値であり、式(18)及び式(19)のように定義される。
Claims (12)
- プロセッサと、前記プロセッサに接続されるメモリと、前記プロセッサに接続される記憶媒体と、前記プロセッサに接続され、他の装置と接続するためのインタフェースとを備え、監視対象の計算機システムから継続して到来するストリームデータを処理するストリームデータ処理サーバであって、
前記ストリームデータは、タイムスタンプが付与された複数のデータを含み、
前記ストリームデータ処理サーバは、予め登録されたクエリにしたがって、スライディングウィンドウを用いて前記ストリームデータから対象となる処理範囲に含まれる前記データを切り出し、前記切り出されたデータに対して分析処理を実行するストリームデータ処理部を備え、
前記ストリームデータ処理部は、
前記切り出されたデータを用いて前記タイムスタンプと前記データの値との対応関係を示す近似式を算出する近似式算出部と、
前記算出された近似式を用いて前記データの予測値を算出し、前記算出された予測値に基づいて前記計算機システムの異常の発生を予測する異常検出部と、を備え、
前記近似式算出部は、
前記切り出されたデータのいずれかのタイムスタンプを時刻の原点に決定し、
前記切り出されたデータの前記タイムスタンプを、前記決定された時刻の原点からの相対的な時刻値に修正し、
修正された前記タイムスタンプと、前記データの値とを用いて前記近似式を算出することを特徴とするストリームデータ処理サーバ。 - 前記近似式算出部は、最小二乗法を用いて前記近似式を算出し、
前記近似式算出部は、
前記切り出されたデータの入力を受け付ける入力部と、
前記入力部が受け付けたデータを格納するデータ格納部と、
前記切り出されたデータに基づいて前記近似式の係数を算出するために用いる統計値を算出する統計値算出部と、
前記算出された統計値を格納する統計値格納部と、
前記統計値を用いて前記近似式の係数を算出する係数算出部と、
を有し、
前記入力部が第1のデータの入力を受け付けた場合に、前記統計値算出部は、前記第1のデータが入力される前に入力された前記データに基づいて算出された第1の統計値と、前記第1のデータとを用いて、第2の統計値を算出することを特徴とする請求項1に記載のストリームデータ処理サーバ。 - 前記入力部は、
前記第1のデータを受け付けた直後に、前記統計値算出部に前記第1のデータを出力し、
前記第1のデータを受け付けてから微小時間経過後に、前記データ格納部に前記第1のデータを出力し、
前記データ格納部は、前記第1のデータが入力される直前に入力された前記データを前記統計値算出部に出力した後に、前記第1のデータを格納し、
前記統計値算出部は、
前記第2の統計値を算出した直後に、前記算出された第2の統計値を前記係数算出部に出力し、
前記第2の統計値を算出してから微小時間経過後に、前記算出された第2の統計値を前記統計値格納部に出力し、
前記統計値格納部は、前記第1の統計値を前記統計値算出部に出力した後に、前記算出された第2の統計値を格納することを特徴とする請求項2に記載のストリームデータ処理サーバ。 - 前記近似式算出部は、
前記処理範囲を変更した場合に、前記変更前の処理範囲に含まれるが、前記変更後の処理範囲に含まれない前記データである消滅データがあるか否かを判定し、
前記消滅データがないと判定された場合には、ztが0である場合の式(7)、式(8)、及び式(9)を用いて前記第2の統計値を算出することを特徴とする請求項5に記載のストリームデータ処理サーバ。 - プロセッサと、前記プロセッサに接続されるメモリと、前記プロセッサに接続される記憶媒体と、前記プロセッサに接続され、他の装置と接続するためのインタフェースとを備え、監視対象の計算機システムから継続して到来するストリームデータを処理する計算機が実行するストリームデータ処理プログラムを記録した記録媒体であって、
前記ストリームデータは、タイムスタンプが付与された複数のデータを含み、
前記メモリは、予め登録されたクエリの定義情報を格納し、
前記ストリームデータ処理プログラムは、
前記クエリの定義情報にしたがって、スライディングウィンドウを用いて前記ストリームデータから対象となる処理範囲に含まれる前記データを切り出す第1の手順と、
前記切り出されたデータを用いて前記タイムスタンプと前記データの値との対応関係を示す近似式を算出する第2の手順と、
前記算出された近似式を用いて前記データの予測値を算出し、前記算出された予測値に基づいて前記計算機システムの異常の発生を予測する第3の手順と、
を前記プロセッサに実行させ、
前記第2の手順は、
前記切り出されたデータのいずれかのタイムスタンプを時刻の原点に決定する手順と、
前記切り出されたデータの前記タイムスタンプを、前記決定された時刻の原点からの相対的な時刻値に修正する手順と、
修正された前記タイムスタンプと、前記データの値とを用いて前記近似式を算出する手順とを含むことを特徴とするストリームデータ処理プログラムを記録した記録媒体。 - 前記ストリームデータ処理プログラムは、
最小二乗法を用いて前記近似式を算出される手順を前記プロセッサに実行させ、
前記データの入力を受け付ける入力手段と、
前記入力部が受け付けたデータを格納するデータ格納手段と、
前記切り出されたデータに基づいて前記近似式の係数を算出するために用いる統計値を算出する統計値算出手段と、
前記算出された統計値を格納する統計値格納手段と、
前記統計値を用いて前記近似式の係数を算出する係数算出手段と、
を有し、
前記第2の手段は、前記入力手段によって第1のデータの入力を受け付けた場合に、前記第1のデータが入力を受け付ける前に入力された全ての前記データに基づいて算出された第1の統計値と前記第1のデータとを用いて、第2の統計値を算出する手順を前記統計値算出手段に実行させる手順を含むことを特徴とする請求項7に記載のストリームデータ処理プログラムを記録した記録媒体。 - 前記第1のデータを受け付けた直後に、前記統計値格納手段に前記第1のデータを出力する手順と、前記第1のデータを受け付けてから微小時間経過後に、前記データ格納手段に前記第1のデータを出力する手順とを、前記入力手段に実行させ、
前記第1のデータが入力される直前に入力された前記データを前記統計値算出手段に出力した後に、前記第1のデータを格納する手順を前記データ格納手段に実行させ、
前記第2の統計値が算出された直後に、前記算出された第2の統計値を前記係数算出手段に出力する手段と、前記第2の統計値が算出されてから微小時間経過後に、前記算出された第2の統計値を前記統計値格納手段に出力させる手順とを、前記統計値算出手段に実行させ、
前記第1の統計値を前記統計値算出手段に出力する手順と、その後、前記算出された第2の統計値を格納する手順とを前記統計値格納手段に実行させることを特徴とする請求項8に記載のストリームデータ処理プログラムを記録した記録媒体。 - 前記第2の手順は、
前記処理範囲を変更した場合に、前記変更前の処理範囲に含まれるが、前記変更後の処理範囲に含まれない前記データである消滅データがあるか否かを判定する手順と、
前記消滅データがないと判定された場合には、ztが0である場合の式(19)、式(20)、及び式(21)を用いて前記第2の統計値を算出する手順と、を前記近似式算出手段に実行させることを特徴とする請求項11に記載のストリームデータ処理プログラムを記録した記録媒体。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2013524548A JP5634607B2 (ja) | 2011-07-20 | 2011-07-20 | ストリームデータ処理サーバ及びストリームデータ処理プログラムを記録した記録媒体 |
PCT/JP2011/066408 WO2013011569A1 (ja) | 2011-07-20 | 2011-07-20 | ストリームデータ処理サーバ及びストリームデータ処理プログラムを記録した記録媒体 |
US14/003,838 US9405795B2 (en) | 2011-07-20 | 2011-07-20 | Stream data processing server and a non-transitory computer-readable storage medium storing a stream data processing program |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2011/066408 WO2013011569A1 (ja) | 2011-07-20 | 2011-07-20 | ストリームデータ処理サーバ及びストリームデータ処理プログラムを記録した記録媒体 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2013011569A1 true WO2013011569A1 (ja) | 2013-01-24 |
Family
ID=47557776
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2011/066408 WO2013011569A1 (ja) | 2011-07-20 | 2011-07-20 | ストリームデータ処理サーバ及びストリームデータ処理プログラムを記録した記録媒体 |
Country Status (3)
Country | Link |
---|---|
US (1) | US9405795B2 (ja) |
JP (1) | JP5634607B2 (ja) |
WO (1) | WO2013011569A1 (ja) |
Families Citing this family (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10102091B2 (en) | 2008-06-04 | 2018-10-16 | Oracle International Corporation | System and method for supporting a testing framework for an event processing system using multiple input event streams |
US10140196B2 (en) * | 2008-06-04 | 2018-11-27 | Oracle International Corporation | System and method for configuring a sliding window for testing an event processing system based on a system time |
JP5634607B2 (ja) * | 2011-07-20 | 2014-12-03 | 株式会社日立製作所 | ストリームデータ処理サーバ及びストリームデータ処理プログラムを記録した記録媒体 |
US10331672B2 (en) * | 2013-06-21 | 2019-06-25 | Hitachi, Ltd. | Stream data processing method with time adjustment |
CN103729444B (zh) * | 2013-12-30 | 2017-07-18 | 清华大学 | 一种基于设备监测数据间潜在关系的异常数据检测方法 |
CN106911589B (zh) * | 2015-12-22 | 2020-04-24 | 阿里巴巴集团控股有限公司 | 一种数据处理方法和设备 |
CN107249096B (zh) * | 2016-06-14 | 2021-02-26 | 杭州海康威视数字技术股份有限公司 | 全景摄像机及其拍摄方法 |
CN106250395B (zh) * | 2016-07-18 | 2019-08-13 | 广西大学 | 一种数据流相似性的连接方法 |
CN110471944A (zh) * | 2018-05-11 | 2019-11-19 | 北京京东尚科信息技术有限公司 | 指标统计方法、系统、设备及存储介质 |
US11689754B2 (en) * | 2019-09-15 | 2023-06-27 | Comscore, Inc. | Systems and methods for predicting viewership and detecting anomalies |
CN115230723A (zh) * | 2022-03-07 | 2022-10-25 | 长城汽车股份有限公司 | 一种车辆预警方法、装置、电子设备与存储介质 |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2010108073A (ja) * | 2008-10-28 | 2010-05-13 | Hitachi Ltd | ストリームデータ処理方法、及びシステム |
JP2010272022A (ja) * | 2009-05-22 | 2010-12-02 | Hitachi Ltd | ストリームデータ処理において逆再生を行うデータ処理システム |
Family Cites Families (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7130971B2 (en) * | 2004-03-30 | 2006-10-31 | Hitachi, Ltd. | Assuring genuineness of data stored on a storage device |
US7873689B2 (en) * | 2004-12-30 | 2011-01-18 | Alcatel-Lucent Usa Inc. | Distributed set-expression cardinality estimation |
JP4687253B2 (ja) * | 2005-06-03 | 2011-05-25 | 株式会社日立製作所 | ストリームデータ処理システムのクエリ処理方法 |
US20060288045A1 (en) * | 2005-06-16 | 2006-12-21 | Digital Fuel Technologies, Inc. | Method for aggregate operations on streaming data |
JP4804233B2 (ja) * | 2006-06-09 | 2011-11-02 | 株式会社日立製作所 | ストリームデータ処理方法 |
JP4933222B2 (ja) * | 2006-11-15 | 2012-05-16 | 株式会社日立製作所 | インデックス処理方法及び計算機システム |
WO2009001827A1 (ja) * | 2007-06-26 | 2008-12-31 | Toshiba Tec Kabushiki Kaisha | 顧客行動管理装置,方法及びプログラム |
JP5465413B2 (ja) | 2008-10-29 | 2014-04-09 | 株式会社日立製作所 | ストリームデータ処理方法、及びそのシステム |
JP5396184B2 (ja) * | 2009-07-31 | 2014-01-22 | 株式会社日立製作所 | 計算機システム及び複数計算機によるストリームデータ分散処理方法 |
US8595194B2 (en) * | 2009-09-15 | 2013-11-26 | At&T Intellectual Property I, L.P. | Forward decay temporal data analysis |
JP4880025B2 (ja) * | 2009-11-26 | 2012-02-22 | 株式会社日立製作所 | ストリームデータ処理方法、ストリームデータ処理プログラム及びストリームデータ処理装置 |
JP5634607B2 (ja) * | 2011-07-20 | 2014-12-03 | 株式会社日立製作所 | ストリームデータ処理サーバ及びストリームデータ処理プログラムを記録した記録媒体 |
-
2011
- 2011-07-20 JP JP2013524548A patent/JP5634607B2/ja not_active Expired - Fee Related
- 2011-07-20 US US14/003,838 patent/US9405795B2/en not_active Expired - Fee Related
- 2011-07-20 WO PCT/JP2011/066408 patent/WO2013011569A1/ja active Application Filing
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2010108073A (ja) * | 2008-10-28 | 2010-05-13 | Hitachi Ltd | ストリームデータ処理方法、及びシステム |
JP2010272022A (ja) * | 2009-05-22 | 2010-12-02 | Hitachi Ltd | ストリームデータ処理において逆再生を行うデータ処理システム |
Non-Patent Citations (1)
Title |
---|
TADASHI MANO: "Ubiquitous Jidai no DB Modeling, Data Administration & Data modeling", SOLUTION IT, vol. 16, no. 2, 1 February 2004 (2004-02-01), pages 44 - 48 * |
Also Published As
Publication number | Publication date |
---|---|
US9405795B2 (en) | 2016-08-02 |
JP5634607B2 (ja) | 2014-12-03 |
JPWO2013011569A1 (ja) | 2015-02-23 |
US20130346441A1 (en) | 2013-12-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP5634607B2 (ja) | ストリームデータ処理サーバ及びストリームデータ処理プログラムを記録した記録媒体 | |
JP4925143B2 (ja) | ストリームデータ処理システム、ストリームデータ処理方法及びストリームデータ処理プログラム | |
US10452983B2 (en) | Determining an anomalous state of a system at a future point in time | |
US10565512B2 (en) | Event analysis apparatus, event analysis method and computer program product | |
US10192050B2 (en) | Methods, systems, apparatus, and storage media for use in detecting anomalous behavior and/or in preventing data loss | |
GB2489831A (en) | Video stream processing | |
US8868993B1 (en) | Data replacement policy | |
WO2015153236A2 (en) | Measuring latency in an interactive application | |
Choi et al. | Inference for discretely observed stochastic kinetic networks with applications to epidemic modeling | |
EP2357562A1 (en) | System for assisting with execution of actions in response to detected events, method for assisting with execution of actions in response to detected events, assisting device, and computer program | |
US20130103373A1 (en) | Online simulation model optimization | |
JP6052278B2 (ja) | 動作判定装置、動作判定システムおよび動作判定方法 | |
CN111045939B (zh) | Weibull分布的故障检测开源软件可靠性建模方法 | |
JP2017037645A (ja) | スマートアラートのためのシステム及び方法 | |
Karlen | Characterizing the spread of CoViD-19 | |
US9455940B2 (en) | Information processing apparatus and information processing method | |
CN104156612A (zh) | 基于粒子滤波的正向与逆向预测误差的故障预报方法 | |
US20190129781A1 (en) | Event investigation assist method and event investigation assist device | |
JP2015513708A (ja) | 高信頼性・高性能のアプリケーションメッセージ配送のためのシステム | |
JP6615470B2 (ja) | 同期制御装置、同期システム及び同期制御方法 | |
Du et al. | On-line control of false discovery rates for multiple datastreams | |
JP2012038135A (ja) | トレンド推移判定装置またはその方法 | |
CN113221096A (zh) | 一种在混沌工程中随机事件相关性分析方法及系统 | |
JP2015187773A (ja) | データ解析装置、データ解析プログラム及びデータ解析方法 | |
Bergström et al. | Bayesian nowcasting with leading indicators applied to COVID-19 fatalities in Sweden |
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: 11869679 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2013524548 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14003838 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 11869679 Country of ref document: EP Kind code of ref document: A1 |