KR101971013B1 - Cloud infra real time analysis system based on big date and the providing method thereof - Google Patents

Abstract

A large data-based cloud infrastructure real-time analysis system according to the present invention includes: a data collection server for collecting cloud operation data of each customer server through a plurality of customer servers and an Analytics as a Service (AaaS) operating in a formed cloud infrastructure; A collection data DB for classifying and storing the cloud operating data collected by the data collection server; A data analysis server for analyzing atypical data of cloud operation data collected by the data collection server to detect an error process and continuously monitoring whether a detected error process is interrupted; And an analysis data DB storing an error process detected from unstructured data in the data analysis server; And the like.
According to the present invention, by analyzing big data in real time in a cloud infrastructure operating environment and continuously monitoring and learning new processes of unstructured data, it is possible to quickly find and cope with the cause of the failure.

Description

[0001] The present invention relates to a real-time analysis system for a cloud infrastructure based on Big Data, and a method for providing the real-

The present invention relates to a large-data-based real-time analysis system for a cloud infrastructure and a method for providing the real-time analysis system. More particularly, the present invention relates to a method and system for analyzing large data in real time in a cloud infrastructure operating environment, And a method for providing the real-time analysis system of the cloud infrastructure.

&Quot; Cloud " computing often refers to the provision of computing resources as a service by multiple computer servers that are networked together in a location remote from the service requesting location. A cloud data center typically refers to the physical arrangement of the servers that make up a particular part of the cloud or cloud. By way of example, servers may be arranged in rooms, groups, rows and racks within a data center. A data center may have one or more " zones " that may include one or more server rooms. Each room may include one or more server rows, and each row may include one or more shelves. Each shelf may include one or more individual server nodes. Servers within a zone, room, shelf and / or column may be arranged in a virtual group based on the physical infrastructure requirements of the data center facility, which may include power, energy, temperature, heat and / or other requirements.

A portion of the server and its resources may be used in accordance with actual or anticipated usage requirements, such as security, quality of service, throughput, throughput and / or other criteria (e.g., (For use by < / RTI > By way of example, a customer's computing workload may be used between multiple physical servers (which may be located in different rows, shelves, groups, or rooms in the data center) or by multiple nodes or resources on the same server Lt; / RTI > Thus, with respect to virtualization, servers can be logically grouped to meet workload requirements.

Today, complex configurations within the cloud data center have made it increasingly difficult to efficiently manage cloud data centers. The main factors contributing to this difficulty are the number of operational data generated by each device and / or service that makes up the data center. Due to the large amount of such data, data center managers often have difficulties in obtaining the health, performance, or even the overall appearance of their data center in real time to check for failures.

Korean Patent Publication No. 2015-0049541

The present invention provides a big data-based real-time cloud infrastructure analysis system and a method for providing real-time analysis of large data based on large data in real time in a cloud infrastructure operating environment to continuously monitor a new process of unstructured data and predict a failure occurrence The purpose.

According to an embodiment of the present invention, there is provided a system for real-time analysis of a cloud infrastructure based on Big Data, comprising: a plurality of client servers and cloud operating data of each client server through an AaaS (Analytics as a Service) A data collection server for collecting the data; A collection data DB for classifying and storing the cloud operating data collected by the data collection server; A data analysis server for analyzing atypical data of cloud operation data collected by the data collection server to detect an error process and continuously monitoring whether a detected error process is interrupted; And an analysis data DB storing an error process detected from unstructured data in the data analysis server; And the like.

In particular, a correspondence management server manages a processing result corresponding to a failure of an error process analyzed by the data analysis server; And a failure pattern data DB storing a processing result of failure detection of an error process of the corresponding management server.

In particular, the cloud operating data is characterized in that the cloud operating data includes fixed data, unstructured data, and status data.

In particular, the format data is characterized by being drive information including a CPU, a memory, a disk, and a network, which are hardware structures.

In particular, the atypical data is characterized by including cloud OS information, hypervisor information, virtual OS information, IT operational performance information, and performance measurement response time.

In particular, the status data includes information on the on / off state of a server, a process, and a DBMS as IT operation status information.

Here, in particular, the data analysis server may include a data real-time analyzing unit for analyzing the collected data, the atypical data and the status data of the collected cloud operating data in real time; A failure pattern DB based analysis unit for comparing the data stored in the failure pattern database with the data type analyzed by the data real time analysis unit to determine a failure; A failure pattern derivation analyzing unit for analyzing a keyword of unstructured data of the data real-time analyzing unit to detect an error process; A failure event generation / notification unit for sending a failure occurrence determined by the failure pattern DB-based analysis unit and an event for occurrence of an error process detected by the failure pattern derivation analysis unit; And a cloud analysis agent synchronization unit for registering the error process information detected by the failure pattern derivation analysis unit.

Here, in particular, the failure pattern derivation analyzing unit is characterized in that the error process is detected by comparing a new keyword text file with a previously registered keyword text (TEXT) file of unstructured data.

According to another aspect of the present invention, there is provided a method for providing a real-time analysis service of a cloud infrastructure based on Big Data, the method comprising: receiving cloud infrastructure operation data formed with a plurality of customer servers; Classifying and analyzing the received operational data into atypical data, fixed data, and status data; Comparing the atypical data analyzed in the analyzing step with a previously stored fault pattern to determine whether the irregular data has failed; Determining whether or not the unstructured data is an error process in comparison with the big data DB if the failure pattern of the atypical data is not detected; And synchronizing the new process to the cloud infrastructure operating AaaS (Analytics as a Service).

Collecting cloud operating data of each customer server through an AaaS (Analytics as a Service) operating in a cloud infrastructure formed with a plurality of customer servers before receiving the cloud infrastructure operating data; And categorizing and storing the cloud operation data collected by the data collection server.

In particular, the big data DB is an analysis data DB including log data, performance data, check data, and event data.

Here, in particular, the step of detecting the new process includes transmitting an event for the detected new process.

In particular, the cloud operating data is characterized in that the cloud operating data includes fixed data, unstructured data, and status data.

In particular, the format data is characterized by being drive information including a CPU, a memory, a disk, and a network, which are hardware configurations of a customer server.

In particular, the atypical data is characterized by including cloud OS information, hypervisor information, virtual OS information, IT operational performance information, and performance measurement response time.

In particular, the status data includes information on the on / off state of a server, a process, and a DBMS as IT operation status information.

In particular, in the step of determining whether or not the fault is present, the analyzed form data is compared with a hardware operation threshold value of the big data DB to determine whether the fault has occurred or not.

In particular, the method includes the step of transmitting a fault occurrence event when the analyzed formulated data exceeds the drive threshold in the determination of the fault, and storing the fault occurrence form data in the big data DB.

In particular, the method includes determining whether the server, the process, and the DBMS, which are the analyzed state data, are turned on / off to determine whether or not the fault has occurred.

In this case, when the state data is determined to be in the off state, the fault occurrence event is transmitted and the fault occurrence state data is stored in the big data DB.

In this case, when the unstructured data is not detected as an error process, the error process failure detection event is transmitted to the big data DB.

In particular, the step of storing the detected new process as unstructured data of the big data DB is performed after the synchronizing.

In particular, when the irregular data is judged to be the fault data in the step of judging whether the irregular data is faulty by comparing the irregular data with the pre-stored fault pattern, the fault occurrence event is transmitted.

According to the present invention, by analyzing big data in real time in a cloud infrastructure operating environment and continuously monitoring and learning new processes of unstructured data, it is possible to quickly find and cope with the cause of the failure.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a diagram schematically illustrating the overall structure of a real-time analysis system for a cloud infrastructure based on Big Data according to an embodiment of the present invention; FIG.
FIG. 2 illustrates a data layer collected in the data collection server of FIG. 1; FIG.
3 is a diagram schematically showing a configuration of the data analysis server of FIG.
FIG. 4 is a flowchart illustrating a method of providing a cloud-based infrastructure analysis service based on a victor according to an embodiment of the present invention; FIG.

While the present invention has been described in connection with certain exemplary embodiments, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, is intended to cover various modifications and similarities. It is to be understood, however, that the invention is not to be limited to the specific embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. In addition, numerals (e.g., first, second, etc.) used in the description of the present invention are merely an identifier for distinguishing one component from another.

Also, in this specification, when an element is referred to as being " connected " or " connected " with another element, the element may be directly connected or directly connected to the other element, It should be understood that, unless an opposite description is present, it may be connected or connected via another element in the middle.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a diagram schematically showing an overall configuration of a real-time analysis system of a cloud infrastructure based on a big data according to an embodiment of the present invention. FIG. 2 is a diagram showing a data layer collected in the data collection server of FIG. And FIG. 3 is a diagram schematically showing the configuration of the data analysis server of FIG.

1, a large data-based real-time cloud infrastructure analysis system according to the present invention includes a plurality of client servers 100, a data collection server 200, a collection data DB 210, a data analysis server 300, an analysis data DB 310, a corresponding management server 400, and a failure pattern data DB 410.

The data collection server 200 collects cloud operation data of each customer server through an analytic service agent (AaaS: not shown) operating in a cloud infrastructure formed with a plurality of customer servers.

Here, the data collection server 200 verifies whether or not the plurality of customer servers 100 is an agent connection registered in advance. For example, it is desirable to verify the host name, IP address, and the like. In other words, it supports a parallel configuration, that is, L4-based hardware load balancing and a store-and-forward method using a task queue so that it can operate even in a large-scale cloud environment.

In addition, the VDI environment supported by the data collection server 200 for the configuration, performance, and fault data collection of the cloud infrastructure includes a Citrix XenDesktop and a VMware View environment. The hypervisor environment includes XEN, KVM, RHEV, VMware vSphere, and Microsoft Hyper-V environments, IBM PowerVM, and HP VM environments. The cloud infrastructure storage includes RDBMS (Relational Database) DB, and NoSQL DB.

As shown in FIG. 2, the cloud operation data collected in the data collection server 200 includes the form data, the unstructured data, and the state data. The formatted data is drive information including a hardware configuration including a CPU, a memory, a disk, and a network, and the atypical data includes cloud OS information, hypervisor information, virtual OS information, IT operational performance information, and performance measurement response time . The status data includes information on the on / off status of the server, the process, and the DBMS as IT operation status information.

The collection data DB 210 classifies and stores environment, event, performance, and failure data of the cloud operating data collected by the data collection server 200.

The data analysis server 300 analyzes the atypical data of the cloud operating data collected by the data collection server 200 to detect an error process and continuously monitors whether the error process is detected.

3, the data analysis server 300 includes a data real-time analysis unit 320, a fault pattern DB-based analysis unit 330, a fault event generation / notification unit 340, a fault pattern derivation analysis unit 350 and a cloud analysis agent synchronization unit 360. [

The data real-time analyzing unit 320 analyzes real-time the form data, unstructured data, and status data types of the collected cloud operating data.

More specifically, the data real-time analyzing unit 320 divides the formatted data of the cloud operating data collected by the data collecting server 200 into a hardware configuration including a CPU, a memory, a disk, and a network driving information, Unstructured data includes cloud OS information, hypervisor information, virtual OS information, IT operational performance information, and performance measurement response time. The state data is classified into IT operation state information including information on the on / off state of the server, the process, and the DBMS.

The data real-time analyzing unit 320 analyzes the cloud operating data transmitted from the customer servers 100 and the analysis service agent (not shown) collected by the data collecting server 200, The data includes at least one of a VDI layer configuration, performance, fault data, VM (Virtual Mashin) layer configuration, performance, fault data, the hypervisor layer configuration, performance and fault data, , Failure data, and the cloud infrastructure storage configuration, performance, and failure data.

The failure pattern DB-based analysis unit 330 compares the data stored in the failure pattern database with data analyzed by the data real-time analysis unit 320 to determine a failure.

More specifically, it is searched whether the irregular data reception contents match the fault pattern DB. The type of the failure pattern DB is managed by a search keyword inputted by an administrator / operator, an important keyword provided by a vendor, a maintenance analysis result or an important keyword through a case occurred in another site, The number of important keywords, and the frequency of each pattern item in unstructured data.

The failure pattern derivation analyzer 350 analyzes the keyword of the atypical data of the data real-time analyzer 320 to detect an error process. Here, the failure pattern derivation detecting unit 350 compares the detection of the error process with a pre-registered keyword text (TEXT) file of irregular data to detect a new keyword text file.

More specifically, a problem process in which the system utilization rate is excessively high, an upper process having a high utilization rate analyzed by the currently operated cloud agent is detected, and system processes such as OS / WAS / DBMS are excluded.

In addition, the same type of text as the unregistered pattern detected in the system in which the failure / problem has occurred, the cloud infrastructure configuration information registered at the time of occurrence of the same infrastructure problem, for example, network, disk, DB connection error, The problem text of the pattern is linked to analyze the association.

The failure event generating / notifying unit 340 transmits an event for the occurrence of the failure determined by the failure pattern DB-based analysis unit 330 and an occurrence of the error process detected by the failure pattern derivation analysis unit. In other words, the trouble type, analysis result, server information (host name, ip address) derived through the above unstructured data analysis are converted into JSON-based standard events in the integrated dashboard and transmitted. That is, the event is transmitted to the integrated dashboard to confirm the administrator of the analyzed problem pattern.

The cloud analysis agent synchronization unit 360 registers the error process information detected by the failure pattern derivation analysis unit 350. [ Here, pattern information is transmitted and synchronized with an AaaS agent currently registered / connected to automate additional monitoring of a new pattern process (error process) not registered in the failure pattern DB-based analyzer 330. That is, the pattern information newly registered in addition to the failure pattern DB information is transmitted to the agent that is newly connected or reconnected and is synchronized.

The analysis data DB 310 stores the error process detected from the unstructured data in the data analysis server 300. Here, the analysis data DB 310 corresponds to the big data DB, and the big data covers all the fixed, irregular and state data. That is, the failure pattern information can be predicted and coped with using the big data including the shape, the irregular shape, and the state.

The correspondence management server 400 manages the processing result corresponding to whether the error process analyzed by the data analysis server 300 has failed or not.

The failure pattern data DB 410 stores the processing result of failure detection of the error process of the corresponding management server 400. [

FIG. 4 is a flowchart illustrating a method of providing a cloud-based infrastructure analysis service based on a large-scale data according to an exemplary embodiment of the present invention.

As shown in FIG. 4, a method for providing real-time analysis service of a cloud infrastructure based on Big Data according to the present invention comprises the steps of: (a) providing a plurality of customer servers 100 with an AaaS (Analytics as a Service) The step of receiving and collecting the cloud infrastructure operational data is performed (S401).

In operation S402, the cloud operating data collected by the data collecting server 200 is analyzed in real time and stored in the collected data DB 210. Here, the received operational data is classified into atypical data, fixed data, and status data and analyzed. And analyzes the collected data, the unstructured data, and the status data type of the cloud operating data in real time.

More specifically, the data real-time analyzing unit 320 divides the formatted data of the cloud operating data collected by the data collecting server 200 into a hardware configuration including a CPU, a memory, a disk, and a network driving information, Unstructured data includes cloud OS information, hypervisor information, virtual OS information, IT operational performance information, and performance measurement response time. The state data is classified into IT operation state information including information on the on / off state of the server, the process, and the DBMS.

In step S403, the irregular data analyzed in step S403 is compared with the previously stored irregular data in step S404. More specifically, it is searched whether the irregular data reception contents match the fault pattern DB. The type of the failure pattern DB is managed by a search keyword inputted by an administrator / operator, an important keyword provided by a vendor, a maintenance analysis result or an important keyword through a case occurred in another site, The number of important keywords, and the frequency of each pattern item in unstructured data.

Next, when the fault pattern of the irregular data is not detected, it is compared with the big data DB to determine whether the irregular data is an error process (S406). That is, the data real-time analyzing unit compares and analyzes the keywords of the big data DB and the unstructured data to determine whether or not it is an error process.

If it is not the error process, it is detected as a new process, and an event for the detected new process is transmitted (S408). Here, the detection of the error process compares the keyword file (TEXT) file of the unstructured data with the previously registered keyword text file (TEXT) to detect the keyword text file of the new process.

More specifically, a problem process in which the system utilization rate is excessively high, an upper process having a high utilization rate analyzed by the currently operated cloud agent is detected, and system processes such as OS / WAS / DBMS are excluded.

In addition, the same type of text as the unregistered pattern detected in the system in which the failure / problem has occurred, the cloud infrastructure configuration information registered at the time of occurrence of the same infrastructure problem, for example, network, disk, DB connection error, The problem text of the pattern is linked to analyze the association.

Then, the detected new process is registered in the monitoring list (S409). That is, since the detected new process may have the possibility of failure data, it is for continuous monitoring.

Then, the step of synchronizing the new process to the cloud infrastructure operating AaaS (Analytics as a Service) is performed (S410).

More specifically, pattern information is transmitted and synchronized to an AaaS agent currently registered / connected to automate additional monitoring of a new pattern process (error process) not registered in the failure pattern DB-based analysis unit. That is, the pattern information newly registered in addition to the failure pattern DB information is transmitted to the agent that is newly connected or reconnected and is synchronized.

After the synchronization, the detected new process is stored as unstructured data of the big data DB (S411). In this case, the text of the detected new process is stored in the big data DB in association with the problem text of the unstructured data pattern and the cloud infrastructure configuration information registered at the time of occurrence of network, disk, DB connection error, etc., It is possible to quickly deal with newly generated failure data through a learning process that is repeatedly performed through data detection.

If it is determined in step S404 that the irregular data has failed, the irregular data is compared with the previously stored irregular data to determine whether the irregular data has failed. If the irregular data is determined to be the irregular data, a fault occurrence event is transmitted in step S405.

If a fault pattern is not detected in the step of detecting the fault pattern of the atypical data (S404), if an irregular data is determined to be an error process as compared with the big data DB (S406), an error process fault detection event is transmitted Is performed (S407).

If the analyzed data is not unstructured data, it is determined whether or not the analyzed data is fixed data (S420). The analyzed data is compared with a hardware operation threshold value of the big data DB (S421) If the formatted data exceeds the drive threshold, it is determined to be failure data, and a failure occurrence event is transmitted (S422). Then, the failure occurrence format data is stored in the big data DB (S423).

If it is determined that the analyzed data is not the formatted data (S430), it is determined whether the analyzed data is the formatted data (S430). That is, it judges whether the server, the process and the DBMS which are the formal data are on / off state and judges whether there is a failure. At this time, if the status data is off in the determination of the failure, the failure is determined to be a failure (S431), the failure occurrence event is transmitted (S432), and the failure occurrence status data is stored in the big data DB .

Finally, it is determined whether the irregular data, the fixed data, and the status data are fault data, and the processed data processing result is stored in the fault pattern database (S412).

Therefore, according to the present invention, in a cloud infrastructure operating environment, large data can be analyzed in real time, a new process of unstructured data can be constantly monitored, and learning through a big data DB can be quickly detected and coped with.

The scope of the present invention is not limited to the above-described embodiments, but may be embodied in various forms of embodiments within the scope of the appended claims. It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the present invention as defined by the appended claims.

100 --- Customer server 200 --- Data collection server
210 --- Collection data DB 300 --- Data analysis server
310 --- Analysis data DB 320 --- Data real time analysis part
330 --- Fault pattern DB-based analysis unit 340 --- Fault event occurrence / notification unit
350 --- Failure pattern derivation analysis part 360 --- Cloud analysis agent synchronization part
400 --- Response management server 410 --- Fault pattern data DB

Claims (23)

In a big data-based cloud infrastructure real-time analysis system,
A data collection server for collecting cloud operating data of each customer server through a plurality of customer servers and an Analytics as a Service (AaaS) operating in a cloud infrastructure formed;
A collection data DB for classifying and storing the cloud operating data collected by the data collection server;
A data analysis server for analyzing atypical data of cloud operation data collected by the data collection server to detect an error process and continuously monitoring whether a detected error process is interrupted; And
An analysis data DB storing an error process detected from unstructured data in the data analysis server; Lt; / RTI >
The data analysis server comprises:
A data real-time analyzer for analyzing the collected data, the unstructured data and the status data of the cloud operating data in real time;
A failure pattern DB based analysis unit for comparing the data stored in the failure pattern database with the data type analyzed by the data real time analysis unit to determine a failure;
A failure pattern derivation analyzing unit for analyzing a keyword of unstructured data of the data real-time analyzing unit to detect an error process;
A failure event generation / notification unit for sending a failure occurrence determined by the failure pattern DB-based analysis unit and an event for occurrence of an error process detected by the failure pattern derivation analysis unit; And
And a cloud analysis agent synchronization unit for registering the error process information detected by the failure pattern derivation analysis unit.
The method according to claim 1,
A correspondence management server for managing a processing result corresponding to whether the error process analyzed by the data analysis server is a failure; And
Further comprising a failure pattern data DB storing a processing result of failure detection of an error process of the corresponding management server.
The method according to claim 1,
The cloud operating data comprises:
Wherein said real-time analysis system comprises at least one of a real-time data, a fixed data, an unstructured data, and status data.
The method of claim 3,
Wherein the formatted data is drive information including a hardware configuration including a CPU, a memory, a disk, and a network.
The method of claim 3,
Wherein the atypical data includes cloud OS information, hypervisor information, virtual OS information, IT operational performance information, and performance measurement response time.
The method of claim 3,
Wherein the state data includes information on the on / off state of the server, the process, and the DBMS as the IT operation state information.
delete The method according to claim 1,
In the failure pattern derivation analysis unit
Wherein the detection of the error process is performed by comparing a new keyword text file with a pre-registered keyword text (TEXT) file of the unstructured data.
A method for providing a real-time analysis service of a cloud infrastructure based on Big Data,
Receiving cloud infrastructure operational data formed with a plurality of customer servers;
Classifying and analyzing the received operational data into atypical data, fixed data, and status data;
Comparing the atypical data analyzed in the analyzing step with a previously stored fault pattern to determine whether the irregular data has failed;
Determining whether or not the unstructured data is an error process in comparison with the big data DB if the failure pattern of the atypical data is not detected; And
And synchronizing the new process to a cloud infrastructure operational AaaS (Analytics as a Service)
Wherein the determination of whether or not the error process is performed is made by comparing a keyword text file of the new process with a previously registered keyword text file of the atypical data, and detecting the keyword text file of the new process.
10. The method of claim 9,
Prior to receiving the cloud infrastructure operational data,
Collecting cloud operating data of each customer server through a plurality of customer servers and an Analytics as a Service (AaaS) operating in a formed cloud infrastructure; And
Further comprising classifying and storing the cloud operation data collected by the data collection server.
10. The method of claim 9,
In the big data DB,
Log data, performance data, check data, and event data.
10. The method of claim 9,
In the step of detecting the new process,
And transmitting an event for the detected new process. ≪ Desc / Clms Page number 19 >
11. The method of claim 10,
The cloud operating data comprises:
Wherein the real-time data includes fixed data, unstructured data, and status data.
14. The method of claim 13,
Wherein the formatted data is drive information including a CPU, a memory, a disk, and a network, which are hardware configurations of a customer server.
14. The method of claim 13,
Wherein the atypical data includes cloud OS information, hypervisor information, virtual OS information, IT operational performance information, and performance measurement response time.
14. The method of claim 13,
Wherein the state data includes information on the on / off state of the server, the process, and the DBMS as the IT operational state information.
10. The method of claim 9,
And comparing the analyzed form data with a hardware operation threshold of the big data DB to determine whether the failure has occurred or not.
18. The method of claim 17,
Wherein the step of transmitting the fault occurrence event and storing the fault occurrence format data in the big data DB is performed when the analyzed formulated data exceeds the drive threshold in the determination of the fault, Delivery method.
10. The method of claim 9,
And determining whether the server, the process, and the DBMS are on / off status of the analyzed state data in the step of determining whether the failure is a failure, .
20. The method of claim 19,
Wherein when the status data is determined to be in the off state, the failure occurrence event is transmitted and the failure status data is stored in the big data DB. Way.
10. The method of claim 9,
And transmitting an error process failure detection event when the unstructured data is determined to be an error process as compared with the big data DB when the failure pattern of the atypical data is not detected. Method of providing analysis.
10. The method of claim 9,
And storing the detected new process as unstructured data of a big data DB after the synchronizing step.
10. The method of claim 9,
Comparing the unstructured data with a previously stored fault pattern to determine whether the unstructured data is faulty; and transmitting the fault occurrence event when the irregular data is determined to be fault data.

Images