JP2013003681A - Service operation management device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To efficiently detect fraudulence of an execution order of services.SOLUTION: A service operation management device includes: a service configuration information management part 111 for storing service execution order definition information defining the execution order of a plurality of services and storing influence range information indicating an influence degree of a failure when the failure occurs in the service; an integrated log storage part 117 for storing logs; and a service failure detection part 115 for inputting a new input log, detecting a previous log of the service executed before the service indicated by the input log from the plurality of logs stored in the integrated log storage part, and determining whether or not the service execution order of the service indicated by the previous log and the service indicated by the input log coincides with the service execution order of the service execution order definition information. The service failure detection part 115 refers to the influence range information stored by the service configuration information storage management part and detects failure occurrence preferentially to the log of the service with a high influence degree of the failure.

Description

  The present invention relates to a service operation management apparatus that detects a failure in a service that is computer software that exchanges messages in the process of providing a function to a user.

  As shown in Patent Document 1, an apparatus that integrates and manages a plurality of types of log data having different recording contents is realized. Using this technology, a service operation management system is built that monitors the ESB log, which is the basic middleware of a service system based on Service Oriented Architecture (SOA), in real time and detects the failure of each service that constitutes the service system. can do. Here, ESB is an abbreviation of “Enterprise Service Bus” and is middleware that performs routing and format conversion necessary for relaying messages between services.

  In a service operation management system comprising a service, an ESB, and a service operation management device according to the prior art, a plurality of services send and receive messages via the ESB to realize functions for users. The service operation management apparatus monitors the ESB log (message log, transaction log which is a record of processing such as message routing performed by the ESB itself) in real time, and detects a service failure. One ESB may be used for each service operation management system, and a plurality of ESBs may coexist due to a corporate merger or partnership.

JP 2010-182194 A

  In the prior art, in order to detect an illegal execution order of services in a complex system, it is necessary to collate all service execution orders that can be executed with the log of the service that is actually being executed. There was a problem that it was difficult.

  In the case of a service operation management system that includes a large number of complex tasks, the real-time nature of log monitoring, such as when the number of service execution order definitions to be matched becomes enormous and many services are executed simultaneously and a large amount of logs are generated in a short time, etc. There was a risk of damage.

  It is an object of the present invention to provide a service operation management apparatus that effectively detects service execution order fraud.

The service operation management apparatus of the present invention
In a service operation management apparatus for detecting a failure in a service provided by computer software executed in a predetermined order,
Service configuration information management unit that stores service execution order definition information that defines the execution order of multiple services, and stores influence range information that indicates the degree of influence of a failure when a failure occurs in response to a service. When,
An integrated log storage unit that stores a plurality of logs as a past log of at least one of a message log that records messages between services and a transaction log that records message relay of the above message;
A new input log to be stored in the integrated log storage unit is input, and a log of a service executed before the service indicated by the input log is detected as a previous log from a plurality of past logs already stored in the integrated log storage unit. The service execution order definition information stored in the service configuration information management unit is retrieved, and the service execution order between the service indicated by the previous log and the service indicated by the input log is stored in the service configuration information management unit. A service failure detection unit that determines whether or not the service execution order matches the service execution order,
With
The service failure detection unit is characterized in that, with reference to the influence range information stored in the service configuration information management unit, the occurrence of a failure is preferentially detected for an input log of a service having a high degree of failure influence.

  According to the present invention, since the service configuration information management unit stores the degree of influence of a failure, it is possible to preferentially detect an illegal service execution order for a case where the degree of influence of a service failure is large.

The block diagram of the service operation management system of Embodiment 1-3. Explanatory drawing of business and service. The figure which shows the hierarchical relationship of the service which the service structure information management part 111 memorize | stores. The figure which shows the service fault influence range information which the service structure information management part 111 memorize | stores. The operation | movement flowchart of the service failure detection part 115. FIG. The operation | movement flowchart of the service failure influence range specific | specification part 112. FIG. The figure which shows the kind of log. The figure which shows the structure of the message which the service part 101 of Embodiment 2 and 3 produces. FIG. 10 is an operation flowchart of the service unit 101 according to the third embodiment. FIG. 10 is an operation flowchart of the attribute query interpretation unit 109 according to the third embodiment. FIG. 10 is an operation flowchart of the attribute query interpretation unit 109 according to the third embodiment. FIG. 3 is a diagram illustrating an example of the appearance of a service unit 101 and a service operation management apparatus 103 according to the first to third embodiments. The figure which shows an example of the hardware resource of the service part 101 in Embodiment 1-3, ESB102, and the service operation management apparatus 103. FIG.

Embodiment 1 FIG.
FIG. 1 is a configuration diagram of a service operation management system 100 using a service operation management apparatus 103 according to Embodiments 1 to 3 of the present invention.

<< Description of Service Operation Management System 100 >>
The service operation management system 100 includes a service unit 101 to be managed, an ESB (Enterprise Service Bus) 102, and a service operation management apparatus 103.

The service unit 101 includes a computer and an application program. The service unit 101 provides a service. A service is achieved by executing an application program on a computer or a dedicated machine. The work is a work that is achieved by sequentially executing services, such as an inventory management work, a reservation work, and a net search work. The business is achieved by executing a plurality of services in a predetermined execution order.
The service unit 101 exchanges messages 104 between services for realizing the functions.

The message 104 has a session ID, a business ID, a service ID, a destination address, a sending address, and a message text.
Here, a session refers to a series of operations and communications from connection or login to disconnection or logoff in a computer system or network communication. The session ID is an identifier assigned to identify a session from connection or login to disconnection or logoff.
The business ID is an identifier assigned to identify a business.
The service ID is an identifier assigned to identify the service.
Further, the message 104 includes attribute information 105 and an attribute query 106 described in Embodiments 2 and 3 to be described later as attribute information.

  The ESB 102 includes a memory device (not shown) that stores a message log 107 and a transaction log 108. The message log 107 records the content of the message 104 as a log. The log records the log time such as year / month / day / hour / minute / second.

  The transaction log 108 is a record of processing such as message routing performed by the ESB 102 itself. The transaction log 108 includes a message 104, a session ID, a business ID, and a service ID.

  As shown in FIG. 1, a plurality of service units 101 exchange messages via the ESB 102 to realize functions for users. One ESB 102 may be used for each service operation management system. However, as shown in FIG. 1, a plurality of ESBs 102 may coexist due to corporate merger or partnership.

<< Configuration of Service Operation Management Device 103 >>
The service operation management apparatus 103 includes a service configuration information management unit 111, a service failure influence range identification unit 112, a collection / accumulation unit 114, a service failure detection unit 115, a log management unit 116, and an integrated log storage unit 117. Furthermore, the service operation management apparatus 103 includes an attribute query interpretation unit 109, an attribute management unit 110, and a service failure definition management unit 113, which will be described in Embodiments 2 and 3 described later.

  The service operation management apparatus 103 receives the log of the ESB 102 (the message log 107 and the transaction log 108 which is a record of processing such as message routing performed by the ESB itself) and monitors it in real time to detect a service failure.

  The service configuration information management unit 111 is a storage device and stores service execution order definition information that defines the execution order of a plurality of services. The service configuration information management unit 111 identifies a service by a business ID and a service ID, and stores a service execution order of a plurality of services by the business ID and the service ID. The service configuration information management unit 111 further stores influence range information indicating the degree of influence of a failure when a failure occurs in the service corresponding to the service.

  The service fault influence range specifying unit 112 searches the service execution order definition information stored in the service configuration information management unit 111 to identify the range of services affected by the fault and is affected by the fault. The service failure information range is stored in the service configuration information management unit 111 as the influence range information.

  The collection / accumulation unit 114 records the transaction log 108 that records message relay between services and the message between services from the ESB 102 (in the case of Embodiment 3 described later, via the attribute query interpretation unit 109). Enter at least one of the message log and log. The log is passed to the log management unit 116.

  The log management unit 116 stores the log collected by the collection / accumulation unit 114 in the integrated log storage unit 117 and delivers the log to the service failure detection unit 115.

  The integrated log storage unit 117 is a storage device, and stores at least one of a transaction log 108 that records message relay between services and a message log 107 that records messages 104 between services as a past log.

  The service failure detection unit 115 inputs a new input log and detects, as a previous log, a past log of a service executed before the service indicated by the input log from a plurality of past logs stored in the integrated log storage unit To do.

  The service failure detection unit 115 searches the service execution order definition information stored in the service configuration information management unit 111, and stores the service execution order of the service indicated by the previous log and the service indicated by the input log in the service configuration information management unit. The occurrence of a service failure is detected by determining whether the service execution order in the service execution order definition information matches.

  Furthermore, the service failure detection unit 115 refers to the influence range information stored in the service configuration information management unit 111, and detects the occurrence of a failure with priority for a service log having a large degree of failure influence.

<< Outline Operation of Service Operation Management Method of Service Operation Management Device 103 >>
The service operation management apparatus 103 according to the present embodiment includes a service configuration information management unit 111 and a service failure influence range specifying unit 112. The service operation management apparatus 103 according to the first embodiment includes characteristic components that are not present in the prior art, such as a service configuration information management unit 111 and a service failure influence range specifying unit 112, and executes the service failure detection unit 115. In this case, priority is given to those having a wide influence range of service failures.

<< Service execution order definition information >>
An example of business and service is shown in FIG.
FIG. 2 shows an example of a service execution order definition in which each business executes a service in the business (1) to (n) which is a set of services. Alphabets (and combinations of subscripts) such as W ₁ and Y ₁ indicate services. The arrow indicates that the service on the left side of the arrow calls and executes the service on the right side of the arrow.

In FIG. 2, the service X is called in the operations (1) to (n), and in order to strictly check whether or not the execution order is invalid when the service X is executed, immediately before the service X, the executed service it is necessary to confirm whether it is one of the services of W ₁ ~W _n. In the case of a service operation management system that includes many tasks with complicated definitions, the service execution order definition to be matched is enormous, and a large number of logs are generated in a short time by executing many services simultaneously. Real-time performance may be impaired.

  In FIG. 3, (1) and (2) are business IDs indicating business. A to G are service IDs indicating services. Business (1) is composed of services A, B, C, and D. Business (2) is composed of services E, B, F, and G. Arrows indicate service calls. In addition, the business using the service B is only (1) and (2). The call relationship between services is called hierarchical information.

  The service configuration information management unit 111 stores the hierarchical information illustrated in FIG. 3 as service execution order definition information (for example, information illustrated in FIG. 3) of a plurality of services using the business ID and the service ID. The storage format and data format of the service execution order definition information are not limited.

≪Effect range information≫
In FIG. 3, assuming that a failure occurs in the service B of the business (1), the influence range of the failure of the service B extends to A, C, and D. The service failure influence range specifying unit 112 calculates the size of the influence range from the service execution order definition information (for example, information of FIG. 3) of the service configuration information management unit 111, and the service configuration information management unit as the influence range information. 111. The influence range information has a degree of influence of a failure when a failure occurs in the service corresponding to the service.

  The influence degree of the failure may be configured to define a fixed value for each service ID, for example. For example, a fixed value is assigned such that service A has 5 points, service B has 10 points, and service C has 7 points, and the fixed values are stored in the service configuration information management unit 111 as influence range information.

  Alternatively, as shown in FIG. 4, when a failure occurs in a certain service, it is assumed that all the services of the business including the service are affected regardless of the service ID of the failed service. The analysis priority order score described in the above section may be assigned to each service, and the total value of the analysis priority order scores of each service may be expressed as a total for each service.

  For example, the analysis priority score gives an initial value of 10 to the service B in which the failure has occurred, and the service A that calls the service B and the services C and D that are called from the service B subtract the analysis priority score for each call. It may be configured to be. For example, the analysis priority score may be calculated by adding instead of subtraction. In FIG. 4, the highest analysis priority score is 10, and the analysis priority score configured so that the analysis priority score is subtracted by 1 for each call is used to perform the tasks (1) and (2) in FIG. This is an example of the influence range information structure when the service fault influence range is calculated.

FIG. 4 is an example of the influence range information stored for the service B of the business (1). In FIG. 4, the influence range information is not a single structure (single value),
1. Structure that holds the detailed impact level for each business,
2. A structure that collectively holds the degree of impact for each business,
3. A structure that collectively holds the impact level for each service,
This is realized as three types of structures.

  For example, if a failure occurs in the service B during the execution of the business (1), the “structure that holds the detailed impact level for each business” includes the degree of influence given to the service B of the business (1) and the business The degree of influence given to the services A, C, and D in (1) is held. If a failure occurs in service B during the execution of business (1), the degree of influence given to service B of business (2) and the degree of influence given to services E, F, and G of business (2) Is retained. In the case of FIG. 4, if a failure occurs in the service B during the execution of the business (1), the service B of the business (2) and the services E and F of the business (2) are not related to the business (2). Since there is no influence on G, the influence degree is kept at zero.

  In the “structure that collectively holds the degree of influence for each business”, the degree of influence for each business is collectively held. In other words, the total of the degree of influence of each service held in the “structure holding the degree of detailed influence by work” is held.

  The “structure that collectively holds the degree of influence for each service” holds the total of the degree of influence of each service that is held in the “structure that holds the degree of detailed influence for each business” for each service.

  From the influence range information of FIG. 4, the influence range of the service B executing the business (1) is 10 points for the service B alone, and 37 points for the business (1) to which the service B belongs, It can be seen that the service (2) to which service B belongs is 0 points.

  From the influence range information of FIG. 4, the influence range of the service B executing the business (2) is 10 points for the service B alone, and 0 points for the business (1) to which the service B belongs. It can be seen that there are 37 points for the service (2) to which the service B belongs.

  The service configuration information management unit 111 stores, as shown in FIG. 4, influence range information indicating the degree of influence of a failure when a failure occurs in the service B corresponding to the service B.

<< Operation of Service Fault Detection Unit 115 Service Execution Order Incorrect Detection Method >>
FIG. 5 shows an operation flowchart of service execution order fraud detection processing in the service failure detection unit 115.

  First, the service failure detection unit 115 inputs from the ESB 102 at least one of a transaction log 108 that records message relay between services and a message log that records messages between services.

  The service failure detection unit 115 receives a log (input log) as an input (S101), and extracts a service ID, a session ID, and a business ID in the input log (S102, S103, S104).

  The service failure detection unit 115 searches the service configuration information management unit 111 and extracts the service failure influence range of the input log (S199). When the service failure influence range of the input log cannot be retrieved from the service configuration information management unit 111, the process proceeds to S105. The service failure detection unit 115 sets a threshold value to be compared with the service failure influence range, and when the service failure influence range of the input log can be searched, the service failure detection unit 115 detects the service having an influence range equal to or greater than the threshold value. When the log is received, the execution order fraud detection process of FIG. 5 is executed with priority over other service failures. When a log of a service having an influence range less than the threshold is received, the input log is stored in a temporary memory, and the execution order fraud detection process shown in FIG. 5 is executed later for the input log.

  In S105, the service failure detection unit 115 retrieves the latest log (referred to as the previous log or previous log) associated with the session ID retrieved in S103 from the integrated log storage unit 117 (S105, S106). The service failure detection unit 115 extracts a log having the same session ID and the latest log time as a previous log.

  If there is a previous log associated with the session ID in S105 and S106, the service failure detection unit 115 performs the processing from S107 onward.

  First, the service failure detection unit 115 searches the service execution order definition information stored in the service configuration information management device 111 by the service ID and the business ID, and executes a service that may be executed next to the service indicated by the previous log. Take out as a candidate service. The service failure detection unit 115 creates a list of service IDs and business IDs of candidate services that may be executed next, which are associated with the previous log extracted in S105 (S107). As will be described later in S114, when the service ID / business ID list is already attached to the previous log extracted in S105, the service ID / business ID list creation operation of the candidate service in S107 can be omitted. .

  Then, it is checked whether the service ID / business ID of the input log extracted in S102 and S103 is included in the list of service IDs / business IDs of candidate services extracted and created in S107 (S109, S110). In S109 and S110, the service failure detection unit 115 determines that the service execution order is not invalid when the service ID and the business ID of the candidate services match, and performs the processing from S113 onward.

  In S108 to S110, when the service ID or the business ID of the candidate service does not match any of the service ID / business ID of the service of the input log extracted in S102 or S103, the service failure detection unit 115 performs the processes of S109 and S110. , Repeatedly executed on the service ID / business ID list of the candidate service extracted and created in S107, and the service ID of the service of the input log extracted in S102 and S103 from the list of the service ID / business ID of the candidate service -Search for the one that matches the business ID. If the service ID / work ID is the last element in the list of service IDs / work IDs of candidate services in S111, the service failure detection unit 115 determines that the service execution order is invalid (S112), and determines in advance. To deal with incorrect service execution order. That is, the service failure detection unit 115 sends service failure information to the service failure influence range specifying unit 112. This service failure information includes a service ID of the service that caused the failure, a service ID, a failure cause ID indicating the cause of the failure, and the like.

  When the service execution order fraud is not detected in the processing of S108 to S110, or when there is no previous log associated with the session ID in S105 and S106, the service failure detection unit 115 performs the processing of S113 and subsequent steps.

  First, the service failure detection unit 115 extracts a list of service IDs and business IDs of candidate services that may be executed next to the service that generated the input log from the service configuration information management unit 111 (S113). And the list of service IDs / service IDs of candidate services extracted in S113 are associated with each other and stored as a past log in the integrated log storage unit 117 through the log management unit 116 (S114), thereby completing the process.

  As described above, the input log stored in the integrated log storage unit 117 will be searched as the past log next time, but the service ID / work ID of the candidate service that may be executed next in the past log. When the list is stored in association with the past log, when the previous log is searched as the previous log, the previous log and the list of service IDs and business IDs that may be executed next can be searched simultaneously. The operation of creating a list of service IDs and business IDs of the candidate services can be omitted.

  Depending on the convenience of the service operation management system 100, the processes of S113 and S114 are also performed on the log in which the service execution order is detected to be invalid, and the service ID of the candidate service is added to the log in which the service execution order is detected in an incorrect manner. A list of business IDs may be associated and stored in the integrated log storage unit 117.

  In FIG. 5, for example, a user ID can be used instead of a session ID depending on the implementation of the service operation management system 100. Further, in the case of a system in which a user executes an individual service instead of a business unit as a set of services, it is possible to omit extraction and comparison of business IDs.

  Alternatively, for example, in S109 and S110, the service ID and the business ID may be compared as a pair, and if either one is different, it may be determined that they do not match.

  As described above, the service failure detection unit 115 is a past log having the same session ID as the session ID included in the input log, immediately before the service indicated by the input log, from the plurality of logs stored in the integrated log storage unit 117. The past log of the executed service is detected as a previous log, the service execution order information stored in the service configuration information management unit 111 is searched, and a service that can be executed after the service indicated by the previous log is set as a candidate service. The service ID of the acquired candidate service and the service ID are determined to match the business ID and service ID of the input log, and the occurrence of a service failure is detected.

<< Operation of Service Fault Influence Range Identification Method of Service Fault Influence Range Identification Unit 112 >>
The service failure influence range specifying unit 112 specifies a service failure influence range. In the following description, a case will be described in which the service failure influence range specifying unit 112 calculates the service failure influence range every time service failure information is received from the service failure detection unit 115.

FIG. 6 shows an operation flowchart of processing of the service failure influence range specifying unit 112.
The service failure influence range specifying unit 112 first receives service failure information from the service failure detection unit 115 (S201). This service failure information includes the business ID, service ID, failure cause ID, and the like of the service that caused the failure.

  Next, the service failure influence range specifying unit 112 extracts a business ID and a service ID for each of the failed services (S202), and extracts the parent service / parent business information of the service from the service configuration information management unit 111. (S203). The parent service information includes the service ID of the parent service, and the parent business information includes the business ID of the parent business.

  The service fault influence range identification unit 112 assembles the impact range information for each business as shown in FIG. 4 for each parent business extracted in S203, and sets the impact range information for each business as part of the service information. Are additionally stored in the service configuration information management unit 111 in correspondence with the failed service (S204).

  In FIG. 6, the service failure influence range specifying unit 112 repeats the process of S204 of FIG. 6 for all parent businesses extracted in S203 (S205), and for all parent services extracted in S203, between services. The hierarchy information is calculated and recorded as intermediate information (S206). Hierarchical information between services refers to the relationship between a service (parent service) calling another service (child service), the relationship between services called by the same parent service (sibling service), and a failure. This information includes the depth of the hierarchy from each service to each service.

  The service failure influence range identification unit 112 recursively executes the processing from S203 to S207 for all parent services extracted in S203 and all parent services / businesses that can follow the parent-child relationship from them (S207). .

  In addition, the service failure influence range specifying unit 112 extracts, from the service configuration information management unit 111, information on child services called by the service for all of the failed services input in S201 (S208), For services, hierarchical information between services is calculated and recorded in intermediate information (S209, S210).

  The service failure influence range specifying unit 112 recursively executes the processing of S208 to S210 for all the child services of each service extracted in S208 and all the child services that can follow the parent-child relationship from them (S211).

  When the above processing is completed, the service failure influence range specifying unit 112 calculates the degree of influence for each business and for each service based on the intermediate information assembled in S206 and S209 (S212, S213). Complete the process.

  A detailed example of the structure of the influence range information is as shown in FIG. In the example of FIG. 4, the detailed influence degree is further calculated for each service and for each service constituting the business, and added to the influence range information structure. The service failure influence range is calculated every time a failure is detected, and the service failure influence range is calculated and stored only for the service in which the failure has occurred.

  When the service failure influence range is calculated and stored only for the service in which the failure has occurred, the service failure influence range cannot be searched for all input logs in S199 of FIG.

  FIG. 6 shows the case where the service failure influence range specifying unit 112 calculates the service failure influence range every time the service failure detection unit 115 detects a failure. The range may be calculated for each log collection / accumulation.

  Further, for example, the service failure influence range specifying unit 112 calculates the service failure influence range for all services every time the service configuration information of the service configuration information management unit 111 is updated, and sets the service failure influence range as a business / service. The service configuration information management unit 111 may associate and store the information in advance. Alternatively, the service failure influence range specifying unit 112 may be configured to temporarily store (cache) the service failure influence range in association with the business / service in the memory device. When the service failure influence range is calculated in advance for all services and stored in the service configuration information management unit 111 in advance, the service failure detection unit 115 determines the service failure influence range for all input logs in S199 of FIG. It is possible to search for all services whether or not to detect preferential service execution order fraud.

With the above configuration and operation, this embodiment has the following effects.
By the operations of the service failure detection unit 115, the service configuration information management unit 111, and the service failure influence range specifying unit 112, it is possible to preferentially detect an invalid service execution order for a wide service failure influence range.

Embodiment 2. FIG.
In the description of the first embodiment, detection of service execution order fraud has been described, but prioritization based on the service fault influence range can also be applied to fault detection other than service execution order fraud.

  The service failures that can be detected from the message log 107 and the transaction log 108 by the service operation management apparatus 103 are shown below.

1. Invalid execution order (in the first embodiment)
Incorrect execution order refers to a case where a service call that deviates from the execution order definition of the service is made.

2. Incorrect timing Incorrect timing refers to a case where a response is not made at a specified time interval (too early or too late) or a response is missing (including a communication error).

3. Service error A service error is an error detected by the service itself. The log contains specific error codes and messages.

4). Invalid value Invalid value is when the returned value is not within the correct range or the format is invalid.

In order to detect these failures, the service operation management apparatus 103 constituting the service operation management system of this embodiment includes a service failure definition management unit 113.
The service failure definition management unit 113 is a storage unit that stores in advance service failure definitions (failure reference values) such as service execution timing, limitation of the range of possible parameter values, and limitation of the range of possible parameter values. .

  In order to detect these failures, the service unit 101 adds an attribute value of the attribute information 105 necessary for detecting the failure to the message 104. As the attribute information 105, a session ID indicating whether it is a series of business, business ID, service execution time, or the like can be added. For example, when the message 104 is XML, the service unit 101 uses a service expression as a conditional expression of the XPath notation indicated by a portion surrounded by <query> to </ query> (within a dotted frame) in FIG. The execution timing, parameter value, service return value, etc. can be defined as attributes.

  The service failure detection unit 115 of the service operation management apparatus 103 determines whether or not the attribute value of the attribute information 105 recorded in the input log violates the abnormal range of the service failure definition information stored by the service failure definition management unit 113. Therefore, a failure can be detected.

  When the service execution timing recorded as the attribute information 105 in the input log does not match the service execution timing, the service failure detection unit 115 determines that a failure has occurred in the service.

  If the parameter value recorded as the attribute information 105 in the input log does not match the possible range of the parameter value, the service failure detection unit 115 determines that a failure has occurred in the service.

  The service failure detection unit 115 has the return value of the service recorded as the attribute information 105 in the input log. If it does not match the possible range of the parameter value, it is determined that a failure has occurred in the service.

  The service failure detection unit 115 is a service with a wide service failure impact range calculated by the service failure impact range specifying unit 112 and violates the service failure definition information stored in the service failure definition management unit 113 Preferentially detect faults. The service failure detection unit 115 gives priority to the failure of the service that violates the service failure definition information stored in the service failure definition management unit 113 regardless of the service failure impact range calculated by the service failure impact range specifying unit 112. You may make it detect to.

  With the above configuration and operation, the service operation management system 100 of this embodiment has the following effects.

  Since the service fault detection unit 115 extracts the attribute information 105 from the message log 107 and collates the value of the attribute information 105 with the service fault definition information stored in the service fault definition management unit 113, the service fault definition management unit 113 It is possible to preferentially detect a service failure that violates the service failure definition stored in the. Further, if the service fault influence range calculated by the service fault influence range specifying unit 112 is also used, it is possible to preferentially detect a fault of an important service.

Embodiment 3 FIG.
In the second embodiment, the attribute extraction from the message needs to be realized depending on the log format of the ESB 102. For this reason, when the service operation management system 100 tries to support the transaction logs 108 of a plurality of ESBs 102 as shown in FIG. Each time the service operation management system 100 is corrected, there is a problem that the flexible and quick service configuration change that is the purpose of the SOA is hindered.

  The ESB 102 processing record has various formats depending on the type of the ESB 102. Examples of these are shown in FIG.

1. Special format ESB log (right side of Fig. 7)
The process of the ESB 102 is recorded in a text-format transaction log 108. Separately, messages between services are stored as files, and the file name and message session ID of the message are recorded in the transaction log 108.

2. General ESB log (left side of Fig. 7)
The transaction log 108 of the ESB 102 includes the entire message between services.

  In the service operation management apparatus according to the prior art, it is necessary to prepare a mechanism (logic) for extracting attributes from the message log or the transaction log 108 according to the ESB 102 log format.

  In FIG. 1, in the service operation management system 100 according to this embodiment, the service unit 101 adds the attribute information 105 to the message 104 and attaches an attribute query 106 for extracting the attribute in the message. In addition, the service operation management apparatus 103 includes an attribute query interpretation unit 109 that interprets the attribute query 106 in the message in the log and extracts the attribute in the message, and an attribute management unit 110 that stores the extracted attribute. Like that.

  An example of the attribute query 106 is shown in the message 104 shown in FIG. In FIG. 8, the portion surrounded by <query> to </ query> (within the dotted frame) indicates the attributes of session (session ID), job (business ID), service (service ID), and time (execution time). This is an attribute query for extraction. In this example, since the message 104 is in the XML format, a query written in a standard XPath notation describing the position of information in the XML document is shown. The attribute may include a value indicating that the message is of high importance. In that case, the service unit 101 adds an attribute query for extracting the attribute of importance.

<< Operation of Service Unit 101 >>
FIG. 9 shows an operation flowchart of processing of the service unit 101. First, the service unit 101 executes a function provided to a user (business service of a company, etc.) (S301), and generates a message 104 based on the result (S302). In generating the message 104, the service unit 101 attaches the attribute information 105 to the message 104 (S303, S304). Further, the service unit 101 generates an attribute query 106 for extracting the attribute information 105 from the message 104, and attaches the attribute query 106 to the message 104 (S305, S306).

  Note that the attribute query 106 may be attached to all of the attribute information 105 or may be attached to a part depending on the convenience in realizing the service operation management system.

  When these processes are completed, the service unit 101 sends the message 104 to the ESB 102 or another service (S307), and ends the process.

<< Operation of Attribute Query Interpreting Unit 109 >>
10 and 11 show operation flowcharts of processing of the attribute query interpretation unit 109 of the service operation management apparatus 103. FIG.

  The service operation management apparatus 103 receives the message log 107 or transaction log 108 of the ESB 102 as input log input, and identifies whether the input log input is the message log 107 or the transaction log 108.

  When the input log is the message log 107, the processing of S501 to S508 is performed.

  When the input log is a transaction log 108 in a general format (left side in FIG. 7), the processing of S410 to S417 is performed.

  When the input log is a special type transaction log 108 (right side in FIG. 7), the processing of S404 to S407 is performed.

  For example, when the message log 107 (and the message 104) is XML and the transaction log 108 is text information having no structure, the identification that the input log is the transaction log 108 is the first line in the message example of FIG. Can be performed by determining whether or not “<? XML version ...?>”. If the first line is “<? XML version...>”, It is determined that the message log 107 (and message 104). If the first line is not “<? XML version ...?>”, It is determined that the transaction log 108 is stored.

  The service operation management apparatus 103 receives the message log 107 or transaction log 108 of the ESB 102 as input log input (S401), and identifies whether the input log input is the message log 107 or not.

  Next, it is identified whether the input log is the message log 107 or the transaction log 108 (S402).

  If the input log is not the message log 107 in S402, the attribute query interpretation unit 109 identifies whether the transaction log 108 includes the message log 107 (XML) as in the general ESB 102 log of FIG. 7 (S403). ). If the message log 107 (XML) is not included (in the case of the transaction log 108 in a special format), the attribute query is interpreted as targeting the transaction log 108, and attribute information not included in the message is extracted (S404). ), The service recorded in the input log is extracted (S405), and service failure detection (including detection of a service failure that is not preferentially executed) is executed by the service failure detection unit 115 (S406). Alternatively, the attribute query interpretation unit 109 performs the same process as the service failure detection unit 115 and detects a service failure.

  The attribute query interpretation unit 109 determines whether the service extracted in S405 is the last service recorded in the input log (S407). If the service is not the last service, the attribute query interpreter 109 determines the next service recorded in the input log. Step S406 is executed for these services. If it is the last service in S407, the attribute query interpretation unit 109 stores the input in the integrated log storage unit 117 by the log storage function of the log management unit 116 (S408), and ends the process.

  In S403, when the transaction log 108 includes the message log 107 (XML) (in the case of a general format transaction log 108), the attribute query interpretation unit 109 extracts the message log (XML) from the input log input. (S410), the attribute query 106 is cut out from each message log (S411). For example, this can be realized by the service unit 101 inserting a reserved word indicating the start point / end point of the attribute query 106 in the message. In step S411, the attribute query interpretation unit 109 determines whether or not the attribute query 106 is the last attribute query 106 in the input log (S412). repeat. If it is the last attribute query being input in S412, the attribute query interpretation unit 109 executes attribute query interpretation (S413). This can be realized, for example, by starting the XPath processing system and inputting the attribute query and message log shown in FIG. 8 to the XPath processing system. Upon receiving this result, the attribute query interpretation unit 109 checks whether or not the attribute includes a value indicating that the message is of high importance (S414). When a value indicating high importance is included here, the service operation management apparatus 103 preferentially executes service failure detection (service execution order illegal detection) by the service failure detection unit 115 (S416). Alternatively, the attribute query interpretation unit 109 performs the same process as the service failure detection unit 115 and detects a service failure.

  When the message does not include a value indicating high importance in the attribute in S414, the attribute query interpretation unit 109 causes the service failure influence range specifying unit 112 to calculate the service failure influence range of the service, and the value is a threshold value. It is determined whether it exceeds (S415). If the threshold value is exceeded, the service operation management apparatus 103 executes S416. That is, the service operation management apparatus 103 preferentially executes service failure detection by the service failure detection unit 115 (S416). Alternatively, the attribute query interpretation unit 109 performs the same process as the service failure detection unit 115 and detects a service failure.

  When the threshold value is not exceeded in S415 or when the process of S416 is completed, the attribute query interpretation unit 109 checks whether or not the message extracted in S410 is the last message body being input (S417). If it is not the last message, the processing of S410 to S417 is repeated. If it is the last message, the attribute query interpretation unit 109 stores the input in the integrated log storage unit 117 by the log storage function of the log management unit 116 (S408). ) End the process.

  In S402, when the input log is the message log 107, the attribute query interpretation unit 109 performs the process of FIG. In FIG. 11, first, the attribute query interpretation unit 109 extracts an attribute query from the input log (S501). This can be realized as the same processing as S411. This process is repeated until the last attribute query in the input log is extracted (S502). Next, the attribute query interpretation unit 109 executes attribute query interpretation for each of the extracted attribute queries (S503). This process can be realized as a process similar to S413. This process is repeated until the last attribute query is reached (S504). Based on this result, the attribute query interpretation unit 109 determines whether the attribute included in the message log 107 includes a value indicating high importance (S505). When a value indicating high importance is included here, the service operation management apparatus 103 preferentially executes service failure detection (S507). If the attribute does not include a value indicating high importance in S505, the attribute query interpretation unit 109 calculates the influence range of the service and determines whether the value exceeds the threshold (S506). If the threshold is exceeded, the service operation management apparatus 103 preferentially executes service failure detection (S507). If the threshold value is not exceeded in S506, the attribute query interpretation unit 109 causes the collection / accumulation unit 114 or the log management unit 116 to store the message log 107 and the attribute until the transaction log 108 including information regarding the log of the message is input. The management unit 110 stores each attribute value (S508). After S507 or S508, the attribute query interpretation unit 109 performs the processing from S409 onward in FIG.

  FIG. 10 shows an example in which failure detection of a highly important service is executed with higher priority than failure detection of a service having a wide service failure influence range. Depending on the implementation convenience, for example, the failure detection of a service having a wide service failure influence range may be executed in preference to the failure detection of a service with high importance.

  Further, the service ID, the business ID, and the session ID in the input log in S102 to S104 in FIG. 5 may be extracted by adding the service ID, the session ID, and the business ID as message attributes to the message 104 and extracting them as attributes. Good. Alternatively, in order to extract the service ID, business ID, and session ID in the input log in S102 to S104 in FIG. 5, an attribute query that extracts the service ID, session ID, and business ID is added to the message, and FIGS. You may implement | achieve by performing an attribute query in such a procedure.

With this configuration and operation, this embodiment has the following effects.
By the operation of the attribute query interpretation unit 109, the attribute information is extracted without depending on the type of log depending on the type of the ESB 102, and the service failure level calculated by the service failure level and the service failure impact range specifying unit 112 indicated by the attribute in the message is calculated. The priority of service failure detection can be determined using either or both of the influence ranges of the parameters as parameters.

As described above, the first to third embodiments have described the service operation management apparatus 103 that detects a failure of a service that is computer software that exchanges messages in the process of providing a function to a user. The service operation management device 103
A service configuration information management unit 111 that manages the definition of the service execution order and the definition of the correspondence between services and services that are a set of services;
Based on the information of the service configuration information management unit, a service failure influence range specifying unit 112 that calculates a range of services that are affected when a failure occurs in a certain service;
A transaction log 108 that is a record of middleware processing that relays messages between services, and a collection / storage unit 114 that collects and stores message logs between services;
A log management unit 116 that accumulates and manages different types of logs and attribute information of the logs in association with each other;
A service failure detection unit 115 for detecting service execution order from the log and detecting service execution order fraud by comparing the service execution order with the service execution order definition of the service configuration information management unit; It is provided with.

  In addition, the service operation management apparatus 103 includes a log text, information for identifying the service that generated the log, information on a session that identifies a series of services to be continuously executed, and the service that has generated the log. A list of identification information of services that may be executed next by the business to which the service belongs is stored in association with the service configuration information management unit 111.

  When the service failure detection unit 115 detects a service execution order fraud from a certain log, the service fault detection unit 115 associated with the log of the service executed immediately before in the same session as the log subject to the fraud detection process is executed. By retrieving a list of service and business identification information that may be executed next by the service from the service configuration information management unit 111 and comparing the list with the service and business identification information of the log subject to fraud detection processing Incorrect service execution order is detected.

In the first to third embodiments, the service operation management system 100 includes:
A service unit 101 which is computer software and hardware for exchanging messages in the process of providing a function to a user;
Middleware that relays messages between services and records individual processes;
The service operation management device described above;
Have
The service operation management apparatus 103 collates the transaction log 108 and the message log with the service execution order definition information stored in the service configuration information management unit 111, and determines the service fault influence range calculated by the service fault influence range specifying unit 112. It is characterized by preferentially detecting an illegal execution order of a wide service.

  Based on the information of the service configuration information management unit 111, the service failure influence range identification unit 112 calculates a range of services that are affected when a failure occurs in a certain service. The service failure influence range specifying unit 112 is configured to obtain hierarchical information between services (a relationship in which a certain service (parent service) calls another service (child service) and a relationship between service groups (sibling services) called by the same parent service). Based on the information), the service failure impact range is specified.

  The service failure influence range identification unit 112 assigns an initial value to a service in which a failure has occurred, and a service that calls the service and a service that is called from the service are configured such that a value is subtracted or added for each call. The service failure influence range is calculated based on the analyzed priority score.

Service operation management device
A service fault definition management unit 113 that stores a service fault definition necessary for detecting a service fault from the aforementioned middleware log and message attributes;
A service failure detection unit 115 that retrieves service failure definition information from the service failure definition management unit 113 and compares the above-described log and message and their attributes to detect a failure that has occurred in the service;
With
The service unit 101 adds attribute information 105, which is additional information necessary for detecting a service failure, to the message 104,
The attribute query interpretation unit 109 extracts attributes from the message log, and the attribute query interpretation unit 109 stores the attribute information, the transaction log 108, the message log, and the attribute value in the service failure definition management unit 113. Collate with the failure definition information and the service execution order definition information stored in the service configuration information management unit 111 to preferentially detect failures of services with a wide range of service failure impacts calculated by the service failure impact range identification unit 112 It is characterized by that.

  The attribute added to the message by the service unit 101 includes information on the importance level of the service. When detecting an illegal service execution order or other service faults, the service importance level is set in addition to the service fault impact range. It is used as a parameter for determining priority of fault detection processing.

  The attribute query interpretation unit 109 extracts an attribute added to a message exchanged between services and an attribute query for extracting the attribute from the message from the message, and also interprets the attribute query and extracts an attribute value. To do.

  The attribute query interpretation unit 109 extracts the attribute from the message by interpreting the attribute query even if the message log is included in the transaction log 108 body in the middleware. Even if the message log is managed as a file different from the transaction log 108, the attribute is extracted from the message without explicitly changing the setting.

  The attribute query interpretation unit 109 detects a service failure according to the priority calculated using the importance of the service indicated by the attribute in the message or the service failure influence range calculated by the service failure influence range specifying unit as a parameter.

FIG. 12 is a diagram illustrating an example of the appearance of the service unit 101 and the service operation management apparatus 103 according to the first to third embodiments.
In FIG. 12, a service unit 101 or a service operation management apparatus 103 includes a system unit 910, a display device 901 having a CRT (Cathode / Ray / Tube) or LCD (liquid crystal) display screen, a keyboard 902 (Key / Board: K / B), hardware resources such as a mouse 903, an FDD 904 (Flexible / Disk / Drive), a compact disk device 905 (CDD), a printer device 906, and a scanner device 907, which are connected by cables and signal lines.

  The system unit 910 is a computer, and is connected to the facsimile machine 932 and the telephone 931 via a cable, and is connected to the Internet 940 via a local area network 942 (LAN) and a gateway 941.

FIG. 13 is a diagram illustrating an example of hardware resources of the service unit 101, the ESB 102, and the service operation management apparatus 103 according to the first to third embodiments.
In FIG. 13, a service unit 101, an ESB 102, and a service operation management apparatus 103 include a CPU 911 (also referred to as a central processing unit, a central processing unit, a processing unit, a processing unit, a microprocessor, a microcomputer, and a processor) that executes a program. I have. The CPU 911 is connected to the ROM 913, the RAM 914, the communication board 915, the display device 901, the keyboard 902, the mouse 903, the FDD 904, the CDD 905, the printer device 906, the scanner device 907, and the magnetic disk device 920 via the bus 912, and the hardware. Control the device. Instead of the magnetic disk device 920, a storage device such as an optical disk device or a memory card read / write device may be used.

  The RAM 914 is an example of a volatile memory. The storage media of the ROM 913, the FDD 904, the CDD 905, and the magnetic disk device 920 are an example of a nonvolatile memory. These are examples of a storage device or a storage unit.

The communication board 915, the keyboard 902, the scanner device 907, the FDD 904, and the like are examples of an input unit and an input device.
Further, the communication board 915, the display device 901, the printer device 906, and the like are examples of an output unit and an output device.

  The communication board 915 is connected to the facsimile machine 932, the telephone 931, the LAN 942, and the like. The communication board 915 is not limited to the LAN 942 and may be connected to the Internet 940, a WAN (wide area network) such as ISDN, or the like. When connected to a WAN such as the Internet 940 or ISDN, the gateway 941 is unnecessary.

  The magnetic disk device 920 stores an operating system 921 (OS), a window system 922, a program group 923, and a file group 924. The programs in the program group 923 are executed by the CPU 911, the operating system 921, and the window system 922.

  The program group 923 stores a program for executing a function described as “˜unit” in the description of the first to third embodiments. The program is read and executed by the CPU 911.

  In the file group 924, information, data, signal values, variable values, and parameters described as “determination results of”, “calculation results of”, and “processing results of” in the description of the first to third embodiments. Are stored as items of “˜file” and “˜database”. The “˜file” and “˜database” are stored in a recording medium such as a disk or a memory. Information, data, signal values, variable values, and parameters stored in a storage medium such as a disk or memory are read out to the main memory or cache memory by the CPU 911 via a read / write circuit, and extracted, searched, referenced, compared, and calculated. Used for CPU operations such as calculation, processing, output, printing, and display. Information, data, signal values, variable values, and parameters are temporarily stored in the main memory, cache memory, and buffer memory during the CPU operations of extraction, search, reference, comparison, operation, calculation, processing, output, printing, and display. Is remembered.

  The arrows in the flowcharts described in the description of the first to third embodiments mainly indicate input / output of data and signals. The data and signal values are the RAM 914 memory, the FDD 904 flexible disk, the CDD 905 compact disk, and the magnetic field. The data is recorded on a recording medium such as a magnetic disk of the disk device 920, another optical disk, a mini disk, and a DVD (Digital Versatile Disk). Data and signals are transmitted online via a bus 912, signal lines, cables, or other transmission media.

  In addition, what has been described as “to part” in the description of the first to third embodiments may be “to circuit”, “to device”, “to device”, and “to means”. It may be “step”, “˜procedure”, “˜processing”. In other words, what has been described as “˜unit” may be realized by firmware stored in the ROM 913. Alternatively, it may be implemented only by software, or only by hardware such as elements, devices, substrates, and wirings, by a combination of software and hardware, or by a combination of firmware. Firmware and software are stored as programs in a recording medium such as a magnetic disk, a flexible disk, an optical disk, a compact disk, a mini disk, and a DVD. The program is read by the CPU 911 and executed by the CPU 911. That is, the program causes the computer to function as the “˜unit” described in the first to third embodiments. Alternatively, the computer executes the procedure and method of “to part” described in the first to third embodiments.

  DESCRIPTION OF SYMBOLS 100 Service operation management system, 101 Service part, 102 ESB, 103 Service operation management apparatus, 104 message, 105 attribute information, 106 attribute query, 107 message log, 108 transaction log, 109 attribute query interpretation part, 110 attribute management part, 111 Service configuration information management unit, 112 Service failure influence range identification unit, 113 Service failure definition management unit, 114 Collection / accumulation unit, 115 Service failure detection unit, 116 Log management unit, 117 Integrated log storage unit, 901 Display device, 902 Keyboard 903 mouse, 904 FDD, 905 compact disc device, 906 printer device, 907 scanner device, 910 system unit, 911 CPU, 912 bus, 913 ROM, 914 RAM, 15 communication board, 920 a magnetic disk device, 921 operating system, 922 Window system, 923 Program group, 924 File group, 931 telephone, 932 a facsimile machine, 940 Internet, 941 Gateway, 942 LAN.

Claims (8)

In a service operation management apparatus for detecting a failure in a service provided by computer software executed in a predetermined order,
Service configuration information management unit that stores service execution order definition information that defines the execution order of multiple services, and stores influence range information that indicates the degree of influence of a failure when a failure occurs in response to a service. When,
An integrated log storage unit that stores a plurality of logs as a past log of at least one of a message log that records messages between services and a transaction log that records message relay of the above message;
A new input log to be stored in the integrated log storage unit is input, and a log of a service executed before the service indicated by the input log is detected as a previous log from a plurality of past logs already stored in the integrated log storage unit. The service execution order definition information stored in the service configuration information management unit is retrieved, and the service execution order between the service indicated by the previous log and the service indicated by the input log is stored in the service configuration information management unit. A service failure detection unit that determines whether or not the service execution order matches the service execution order,
With
The service failure detection unit refers to the influence range information stored in the service configuration information management unit, and preferentially detects the occurrence of a failure with respect to an input log of a service having a large influence degree of the failure. Operation management device. The log has a session ID, a business ID, and a service ID.
The service configuration information management unit defines a service by a business ID and a service ID, stores a service execution order of a plurality of services by the business ID and the service ID,
The service failure detection unit is a past log having the same session ID as the session ID of the input log from a plurality of past logs already stored in the integrated log storage unit, and the service executed immediately before the service indicated by the input log The previous service log is detected as the previous log, and the service execution order is searched from the service execution order definition information stored in the service configuration information management unit, and a service that can be executed after the service indicated by the previous log is selected as a candidate service. 2. A service failure occurrence is detected by determining whether the acquired business ID and service ID of the candidate service match the business ID and service ID of the input log. Service operation management device. The service operation management device
By retrieving the service execution order from the service execution order definition information stored in the service configuration information management unit, the range of other services affected by the failure is identified from the failed service where the failure occurred, and affected by the failure. 3. A service failure influence range specifying unit for storing a failure influence range of another service as a failure occurrence service influence range information in association with a failure occurrence service and storing it in a service configuration information management unit. The service operation management device described.   4. The service operation management apparatus according to claim 1, wherein the service fault influence range specifying unit specifies a service fault influence range based on the hierarchy information using the call relationship between services as hierarchy information. .   The service fault influence range identification unit assigns an initial value to the fault service, and analyzes the value that is subtracted or added for each call for services that call the fault service and services that are called from the fault service The service operation management apparatus according to claim 4, wherein the service operation management apparatus is assigned as a priority score, and calculates a failure influence range of the service based on the analysis priority score. The service operation management device
A service fault definition management unit for storing service fault definition information in which an abnormal range of attribute values of attribute information included in a service message is defined as a fault reference value;
The message has attribute information including the attribute value of the service,
The service fault detection unit refers to the service fault definition information stored in the service fault definition management unit, compares the fault reference value defined by the service fault definition information with the attribute value of the attribute information included in the log message. 6. The service operation management apparatus according to claim 1, wherein occurrence of a service failure is detected. The message has attribute information including an attribute value indicating the importance of the service,
The service failure detection unit preferentially detects the occurrence of a service failure of a message having attribute information having a high importance level with reference to the importance level indicated by the attribute information included in the log message. -Service operation management apparatus in any one of -6. The message has an attribute query for retrieving attribute information from the message,
The service operation management device
Execute an attribute query included in the log message and extract attribute information from the log message;
An attribute management unit for storing attribute information extracted by the attribute query interpretation unit,
8. The service operation management apparatus according to claim 6, wherein the service failure detection unit determines failure detection with reference to attribute information stored in the attribute management unit.

Images