JP5821471B2 - Information processing apparatus, process monitoring method, process monitoring program, and recording medium - Google Patents

Description

  The present invention relates to an information processing apparatus, a process monitoring method, a process monitoring program, and a recording medium.

  In a Java (registered trademark) virtual machine, a process may end abnormally. As a cause of abnormal termination of the process, for example, there is a shortage of heap area. The heap area is one of the memory areas used by the process, and an upper limit is set for the size thereof. If the process terminates abnormally, the business application program (hereinafter referred to as business application) cannot be executed in the process, and the business application cannot be continuously executed.

  Therefore, the possibility that the process ends abnormally before the process ends abnormally is detected, and a warning is output to the operator, for example. For example, when the usage amount of the heap area exceeds a threshold value, a warning message is output. Thereby, the abnormal termination of the process is avoided and the execution of the business application is continued.

  It has also been proposed to detect abnormalities by monitoring and comparing computer hardware information. For example, the request stop node selection means selects a node that stops distribution of requests from a plurality of fault monitoring nodes constituting the cluster, and the request stop means stops distribution of the request to the selected node, and the rare node The calculation means obtains the detailed information of the selected node from the detailed information storage unit, and using this detailed information, the resource usage status is different from the other faulty monitoring nodes among the plurality of faulty monitoring nodes. It is proposed that the cluster failure estimation means detects a node in an abnormal state using the calculation result calculated by the rare node calculation means.

  In addition, a plurality of controllers that repeatedly capture a state of a control target and output a signal determined by executing a task provided in advance to the control target are provided, and the plurality of controllers are connected via a network. In the control system, a system simulator that stores the activation relationship of the task to be controlled, and identifies a task that is about to be executed in the near future based on information on the activation relationship of the task and information on the task currently being executed by the controller, It has been proposed to include a task broker that determines a controller that executes the specified task by the system simulator from the plurality of controllers.

JP 2007-122330 A Japanese Patent No. 3787359

  For example, when there is a process in which some kind of abnormality such as an application failure exists, the usage amount of the heap area of the process in which the abnormality has occurred may not exceed the threshold. In this case, the possibility of abnormal termination of the process cannot be detected even if the heap area usage threshold is used.

  When an abnormality is detected by monitoring and comparing computer hardware information, the hardware information is compared on a computer basis. For this reason, an abnormality specific to a business application cannot be detected.

  An object of one aspect of the present invention is to detect the possibility of abnormal termination of a process due to an application.

  The disclosed information processing apparatus includes: a collection unit that collects characteristic information representing process characteristics for each of a plurality of processes that execute processing based on a plurality of application programs; and characteristic information related to a certain process among the plurality of processes Includes a determination unit that determines that an abnormality has occurred in a certain process when the characteristic is different from any characteristic information of any of the plurality of processes excluding the certain process.

  According to one aspect of the information processing apparatus disclosed, it is possible to detect a possibility that a process will end abnormally due to an application.

It is a figure which shows an example of an information processing system. It is a figure which shows an example of information processing apparatus. It is explanatory drawing of a process monitoring process. It is a figure which shows an example of a hardware structure of information processing apparatus. It is a process monitoring process flow. It is a process monitoring process flow. It is a process investigation process flow.

  Depending on the type of abnormality in the process, the threshold value for detecting the abnormality itself may not be determined. In this case, it is impossible to detect the possibility of abnormal termination of the process using the threshold value for abnormality detection.

  For example, in HTTP (Hypertext Transfer Protocol), the status code 200 (OK) is defined as a code indicating that the “request was successful” from the client. When the status code is “200”, information corresponding to the request is returned together with a response to the request from the client, and this is a case of normal return. However, in practice, a status code other than “200” may occur. In other words, there may be a case where the normal return is not achieved.

  Therefore, when a status code other than “200” occurs, it is conceivable to disconnect the computer in which the abnormality has occurred from the information processing system. However, in this case, although the computer itself has no abnormality, the computer is frequently disconnected, resulting in an obstacle to the execution of actual work. Therefore, even if a status code other than “200” is generated, it is not appropriate to determine that an abnormality has occurred in the computer.

  On the other hand, when there is a tendency that many status codes other than “200” tend to occur, there is a possibility that an abnormality such as destruction of business application data has occurred, not a computer abnormality.

  However, the frequency of occurrence of status codes other than “200” usually differs for each business application and takes a value specific to each business application. In other words, it is impossible to calculate the frequency of occurrence of status codes other than “200” as a value common to various business applications. Therefore, in various business applications, it is not possible to determine that it is abnormal if a status code other than “200” is generated a certain number of times. For this reason, an appropriate threshold value for detecting an abnormality cannot be determined using a status code other than “200”.

  Further, depending on the type of abnormality in the process, even if a threshold value for abnormality detection is set, the possibility of abnormal termination of the process cannot be detected.

  For example, if requests are evenly distributed to a plurality of business processes and the amount of resources used depending on the requests does not vary greatly, each of the plurality of business processes should consume the same amount of resources. Therefore, in this case, it is conceivable to set a threshold value for detecting an abnormality using the resource consumption.

  However, as a result of some abnormality such as an application failure, only a specific business process may consume a large amount of resources. In some cases, only a specific business process consumes a large amount of resources, but the amount of resource consumption does not exceed a threshold for detecting an abnormality. In this case, an abnormality cannot be detected even if a threshold value for detecting an abnormality is set.

  As described above, when a computer in which an abnormality has occurred is disconnected from the information processing system, for example, an alternative computer is prepared separately from a computer that executes business, such as a cluster system. In this case, the computer that was executing the business application in which the abnormality occurred is stopped, and the execution of the business application by the alternative computer is started. Thereby, the execution of the business application is continued. However, since the computer in which the abnormality has occurred is stopped, the state of the process or the business application when the abnormality has occurred cannot be maintained, and the cause of the abnormality cannot be elucidated.

  The disclosed information processing apparatus detects the possibility of abnormal termination of a process due to an application without using a threshold value.

  FIG. 1 is a diagram illustrating an example of an information processing system.

  The information processing system is a server client system, for example, and includes an application server 1, a client 8, and a network 9. The application server 1 is a server device that executes the business application 4, in other words, executes business processing. The network 9 is a network such as the Internet, for example. The client 8 is a client device that transmits a business request to the application server 1 and requests execution of the business application 4.

  For example, the application server 1 receives a business request from the client 8 via the network 9. Then, the application server 1 executes the business application 4 specified in the received business request. Thereafter, the application server 1 transmits a response to the business request from the client 8 and information corresponding to the business request, in other words, the execution result of the business application 4 to the client 8 via the network 9.

  The application server 1 includes a distribution process 2, a plurality of business processes 3, a business application 4, a plurality of alternative processes 5, a business application 6, and a monitoring process 7.

  The distribution process 2 is a distribution unit that executes distribution processing, and distributes business requests from the clients 8 to a plurality of business processes 3. In other words, the distribution process 2 sends a business request to the business process 3 and requests the business process 3 to execute the business request.

  Specifically, the distribution process 2 receives a business request from the client 8 when receiving a notification of the completion of the activation of the business process 3 from the monitoring process 7, in other words, accepts it. Then, the distribution process 2 determines which of the plurality of business processes 3 the received business request is distributed to. Thereafter, the distribution process 2 distributes the business request to the business process 3 based on the determination, in other words, requests the business process 3 to execute the business request. For example, the business requests are equally distributed to each of the plurality of business processes 3.

  Further, the distribution process 2 distributes the business request received from the client 8 to the alternative process 5 when the notification of the start completion of the alternative process 5 is received from the monitoring process 7. In other words, when the distribution process 2 receives the notification of the completion of the activation of the alternative process 5 from the monitoring process 7, the distribution process 2 sends the business request received from the client 8 to the business process 3 that has distributed the business request up to that point. Avoid sorting. As a result, the business process 3 is switched to the alternative process 5.

  The distribution process 2 is activated, for example, by executing a predetermined activation command for the distribution process 2 prior to the start of business. Specifically, for example, the administrator of the application server 1 inputs a start command for the distribution process 2 from the input device of the application server 1, thereby causing the application server 1 to execute the start command for the distribution process 2.

  The business process 3 is a business process execution unit that actually executes a business process, and executes a business request from the client 8 distributed by the distribution process 2. In other words, the business process 3 executes processing instructed in the received business request.

  Specifically, when the business process 3 receives a business request from the distribution process 2, the business process 3 calls the business application 4 that executes the received business request, and causes the called business application 4 to execute the process specified in the business request. . In other words, the business process 3 executes processing based on the business application 4. When a plurality of business requests are distributed, the business process 3 causes the business application 4 to process the business requests for each distributed business request. As a result, the business process 3 executes the processing requested by the client 8. The business application 4 is an application program executed in the business process 3.

  The business process 3 is started, for example, when the monitoring process 7 executes a predetermined start command for the business process 3. For example, when the monitoring process 7 is started, the monitoring process 7 executes a start command for the business process 3. At this time, a plurality of business processes 3 are started and operated in parallel. Thereby, the load can be distributed and the application server 1 can be operated stably.

  The alternative process 5 is a process that executes processing in place of the business process 3 when an abnormality occurs in the business process 3. Therefore, the alternative process 5 has the same function as the business process 3.

  The alternative process 5 is a business process execution unit that actually executes a business process, and executes a business request from the client 8 distributed by the distribution process 2. Specifically, when receiving the business request from the distribution process 2, the alternative process 5 calls the business application 6 that executes the received business request and causes the called business application 6 to execute the process specified in the business request. . When a plurality of business requests are distributed, the alternative process 5 causes the business application 6 to process the business requests for each distributed business request. Thereby, the alternative process 5 executes the process requested by the client 8.

  The business application 6 is an application program executed in the alternative process 5. The business application 6 is actually an application program similar to the business application 4.

  The alternative process 5 is activated, for example, when the monitoring process 7 executes a predetermined activation command for the alternative process 5. For example, when an abnormality occurs in the business process 3, the monitoring process 7 executes a startup command for the alternative process 5. Note that the monitoring process 7 may execute the activation command of the alternative process 5 immediately after the monitoring process 7 is activated, for example.

  The number of activated alternative processes 5 is the same as the number of activated business processes 3, for example. As a result, when an abnormality occurs in the business process 3, not only the business process 3 in which the abnormality has occurred, but all the business processes 3 can be switched to the alternative process 5. Note that the number of activated alternative processes 5 may be smaller than the number of activated business processes 3.

  The monitoring process 7 is an activation unit that activates a process, and activates the business process 3 and the alternative process 5. As described above, when the monitoring process 7 is started by the start command of the monitoring process 7, the start process command of the business process 3 is executed a plurality of times to start a plurality of business processes 3. In addition, when an abnormality occurs in the business process 3, the monitoring process 7 starts the plurality of alternative processes 5 by executing the startup command for the alternative process 5 a plurality of times.

  The monitoring process 7 is a monitoring unit that executes a monitoring process, and is a process that monitors all the processes started by itself, in other words, the business process 3 and the alternative process 5. The monitoring process 7 monitors the execution of the business by monitoring the business process 3 and the alternative process 5 that execute the business, and determining whether or not an abnormality has occurred in the business process 3. When the business process 3 is switched to the alternative process 5, the monitoring process 7 monitors the execution of the business by determining whether an abnormality has occurred in the alternative process 5, which is a new business process. The monitoring process 7 will be described later with reference to FIG.

  The monitoring process 7 is activated, for example, by executing a predetermined activation command for the monitoring process 7 prior to starting the business. Specifically, for example, when the administrator of the application server 1 inputs a start command for the monitoring process 7 from the input device of the application server 1, the application server 1 executes the start command for the monitoring process 7.

  FIG. 2 is a diagram illustrating an example of an information processing apparatus. FIG. 3 is an explanatory diagram of the process monitoring process. In FIG. 2, the alternative process 5 is not shown.

  The monitoring process 7 activates a plurality of business processes 3, for example, business process # 1 and business process # 2. The number of business processes 3 to be activated is not limited to two. Also, the monitoring process 7 activates a plurality of alternative processes 5, for example, the alternative process # 1 and the alternative process # 2. As described above, the number of activated alternative processes 5 is the same as the number of activated business processes 3.

  Thereafter, the monitoring process 7 generates business process management information 71 shown in FIG. 3A and transmits the generated business process management information 71 to the distribution process 2. Actually, the business process management information 71 is added to the notification of the start completion of the business process 3. For example, when the monitoring process 7 receives notification of activation completion from all activated business processes 3, the monitoring process 7 transmits a notification of activation completion of the business process 3 to the distribution process 2.

  As shown in FIG. 3A, the business process management information 71 is information representing a business process 3 that actually executes a business. In other words, the business process 3 is a process that can distribute a business request from the distribution process 2. The business process management information 71 is generated by the monitoring process 7 every time the business process 3 is activated.

  The business process management information 71 includes a process ID and a port number for each process number. The process number is identification information that uniquely defines a process in which a business request is distributed and an actual business is executed, in other words, the business process 3 among the activated processes. The process ID is identification information that uniquely defines the activated process. The activated processes include the business process 3, the alternative process 5, and all other processes. The port number is identification information that uniquely defines a port, for example, a logical port, used by the process with the corresponding process ID in inter-process communication.

  When the distribution process 2 receives the notification of the start completion of the business process 3 and the business process management information 71 from the monitoring process 7, the distribution process 2 distributes the business request received from the client 8 to the business process 3 based on the business process management information 71. Divide. In response to this, the business process 3 to which the business request has been distributed executes the business requested in the business request by the business application.

  On the other hand, the monitoring process 7 monitors the business process 3 after sending a notice of the completion of the activation of the business process 3 to the distribution process 2. Specifically, the monitoring process 7 collects business process information from all the business processes 3 periodically at a predetermined cycle. Collection of business process information is executed by the collection unit 73. The business process information is characteristic information that represents the characteristics of each business process 3. The collection unit 73 collects characteristic information for each of the plurality of business processes 3, in other words, for each of the plurality of business processes 3.

  For example, the business process information 31 shown in FIG. 3B is collected from the business process # 1, and the business process information 32 shown in FIG. 3C is collected from the business process # 2. As shown in FIG. 2, information collected from the business process # 1 is also referred to as business process information # 1, and information collected from the business process # 2 is also referred to as business process information # 2.

  As shown in FIG. 3B, the business process information 31 is characteristic information representing the characteristics of the business process # 1. The business process information 31 is generated by the business process # 1 and updated sequentially.

  The business process information 31 includes a process ID, memory usage, response time, and “number of times other than 200”. The process ID is its own process ID, in other words, the process ID of the business process # 1. The memory usage is the amount of memory used by the business process # 1 at that time. The response time is the time from when the business process # 1 receives the business request distributed to itself until the response to the business request is returned. “Number of times other than 200” is the number of times a status code other than the status code “200” in HTTP has occurred in the business process # 1. HTTP is a protocol used when a plurality of business processes 3 communicate with, for example, the client 8 or the distribution process 2.

  The business process information 32 is characteristic information representing the characteristics of the business process # 2, as shown in FIG. The business process information 32 is generated by the business process # 2 and updated sequentially.

  The business process information 32 includes a process ID, memory usage, response time, and “number of times other than 200”. The process ID is its own process ID, in other words, the process ID of the business process # 2. The memory usage is the amount of memory used by the business process # 2 at that time. The response time is the time from when the business process # 2 receives the business request distributed to itself until the response to the business request is returned. “Number of times other than 200” is the number of times a status code other than “200” has occurred in the business process # 2.

  In the business process information 31 and 32, memory usage, response time, and “number of times other than 200” are characteristic information. Each of the business process information 31 and 32 includes a plurality of types of characteristic information. The characteristic information is information about a predetermined specific process or information about a predetermined specific resource. Specifically, the characteristic information is information indicating the number of times the result of processing executed by the process is a predetermined processing result, or information on the usage amount of a predetermined resource used by the process.

  The memory usage includes, for example, a heap area, a C heap area, and a permanent generation area in a Java virtual machine. As the response time, in addition to the response time for each business process 3, the response time for each business application 4 and the response time for each business request may be used. As the “number of times other than 200”, the number of times a status code other than the status code of the 200th generation is generated in the business process # 1 may be used, and status codes of the 300th and 400th generations are generated. The number of times may be used.

  The characteristic information is not limited to data related to resources used by the business process 3, such as memory usage, response time, and “number of times other than 200”. For example, unique information of the business application 4, database connection, thread pool, transaction information, file descriptor, handle, CPU usage rate, and the like may be used as the characteristic information.

  The unique information of the business application 4 used as the characteristic information is, for example, a state where the business application 4 is prohibited. The database connection used as the characteristic information is, for example, the number of sockets when the business application 4 accesses the database. The thread pool used as the characteristic information is, for example, the number of threads pooled by the business process 3. The transaction information used as the characteristic information is, for example, the number and time of transactions in the business process 3. The file descriptors and handles used as characteristic information are the numbers that the business process 3 has. The CPU usage rate used as the characteristic information is, for example, a rate at which the business process 3 uses the CPU.

  For example, the collection unit 73 of the monitoring process 7 requests each of the plurality of business processes 3 to transmit characteristic information. In response to this, each of the plurality of business processes 3 transmits characteristic information to the collection unit 73 of the monitoring process 7. Thereby, characteristic information is collected. The characteristic information may be collected by referring to the characteristic information written in the storage device such as the shared memory by the business process 3 by the collection unit 73 of the monitoring process 7.

  When the monitoring process 7 collects the business process information from all the business processes 3, the monitoring process 7 compares the collected business process information. The comparison of the business process information is executed by the comparison unit 74 and is executed every time the business process information is collected from all the business processes 3.

  For example, the comparison unit 74 uses the business process information collected from each of the plurality of business processes 3 as the first business process 3 and the second business process 3 that is a business process other than the first business process 3. Divide into sets. Then, the comparison unit 74 compares the characteristic information of the first business process 3 with the characteristics indicated by the set of characteristic information of the second business process 3.

  The first business process 3 is a certain business process 3. The first business process 3 is acquired by sequentially extracting one business process 3 from a plurality of business processes 3 that have collected business process information.

  As a result of the comparison in the comparison unit 74, the monitoring process 7 determines that the first business process 3 is different from the characteristic information of the first business process 3 and the characteristic indicated by the set of characteristic information of the second business process 3. Judge that an abnormality has occurred. The determination of whether or not an abnormality has occurred is performed by the determination unit 75 and is performed each time the business process information is compared. The monitoring unit 7, in other words, the determination unit 75 has the characteristic information related to a certain business process 3 among the plurality of business processes 3, and the remaining business processes excluding the certain business process 3 among the plurality of business processes 3. 3, it is determined that an abnormality has occurred in the certain business process 3.

  A criterion for determining whether or not the characteristic information of the first business process 3 is different from the characteristic indicated by the set of characteristic information of the second business process 3 is determined empirically for each type of characteristic information.

  For example, it is assumed that the business process # 1 is the first business process 3 and the business process information # 1 in FIG. 3B is the characteristic information of the first business process 3. Further, it is assumed that the business process # 2 is the second business process 3, and the business process information # 2 in FIG. 3C is the characteristic information of the second business process 3. Furthermore, it is assumed that there are a plurality of business processes 3 having the same business process information as the business process # 2. Therefore, the set of characteristic information of the second business process 3 is a set of business process information # 2 in FIG.

  In this case, the characteristic information of the business process # 1, which is the first business process 3, is “5” when focusing on “number of times other than 200”. On the other hand, the characteristic indicated by the set of characteristic information of the plurality of business processes # 2 as the second business process 3 is “1” when focusing on “number of times other than 200”. Accordingly, only the characteristic information of the business process # 1, which is the first business process 3, has a value different from the characteristic information of the other business processes 3. As a result, the characteristic information of the business process # 1 that is the first business process 3 is different from the characteristic indicated by the set of characteristic information of the plurality of business processes # 2 that are the second business process 3. Therefore, it is determined that an abnormality has occurred in the business process # 1, which is the first business process 3. The number of business processes 3 that are determined to have an abnormality is not limited to one.

  Further, for example, with respect to the memory usage, for example, only the characteristic information of the first business process 3 is increasing, and the characteristics indicated by the set of characteristic information of the second business process 3 are not increasing, in other words, When the value decreases or maintains a substantially constant value, it may be determined that an abnormality has occurred in the first business process 3. On the contrary, only the characteristic information of the first business process 3 is decreasing, and the characteristic indicated by the set of characteristic information of the second business process 3 is not decreasing. In other words, an increase or almost constant value is set. When maintaining, it may be determined that an abnormality has occurred in the first business process 3. Also in this case, the number of business processes 3 that are determined to have an abnormality is not limited to one.

  For example, the response time may be determined in the same manner as the memory usage, for example.

  Further, as described above, each of the business process information 31 and 32 includes a plurality of types of characteristic information. Therefore, for one type of characteristic information, when the characteristic information of the first business process 3 and the characteristic indicated by the set of characteristic information of the second business process 3 are different, the characteristic information of the first business process 3 It may be determined that the characteristic indicated by the set of characteristic information of the second business process 3 is different. Conversely, for each of a plurality of types of characteristic information, it is determined whether the characteristic information of the first business process 3 and the characteristic indicated by the set of characteristic information of the second business process 3 are different. Thus, it may be determined whether or not the characteristic information of the first business process 3 and the characteristic indicated by the set of characteristic information of the second business process 3 are different.

  As described above, it is possible to detect the occurrence of an abnormality in the business process 3 and, in fact, a sign of the occurrence of the abnormality, without using a threshold for abnormality detection. Further, it is possible to detect an abnormality of an application for which a reasonable threshold value for abnormality detection cannot be set. As a result, it is possible to avoid the business process 3 from being abnormally terminated and the business being stopped.

  If the monitoring process 7 determines that an abnormality has occurred in the business process 3, an abnormality has occurred in the business process 3 by, for example, outputting a message to a log or displaying it on a display device to the administrator of the application server 1. Warn you. Note that the administrator of the application server 1 may be notified of the business process 3 in which it is determined that an abnormality has occurred along with or separately from the warning.

  When the monitoring process 7 determines that an abnormality has occurred in the business process 3, the monitoring process 7 activates a plurality of alternative processes 5, for example, the alternative process # 1 and the alternative process # 2. As described above, the number of activated alternative processes 5 is the same as the number of activated business processes 3. Therefore, the number of the alternative processes 5 to be started is not limited to two.

  Thereafter, the monitoring process 7 generates the alternative process management information 72 shown in FIG. 3D and transmits the generated alternative process management information 72 to the distribution process 2. Actually, the alternative process management information 72 is added to the notification of the completion of activation of the alternative process 5. For example, when the monitoring process 7 receives notification of activation completion from all the activated alternative processes 5, the monitoring process 7 transmits the activation completion notification of the alternative process 5 to the distribution process 2.

  As shown in FIG. 3D, the alternative process management information 72 is information representing the alternative process 5 that actually executes a job. The alternative process management information 72 is generated by the monitoring process 7 every time the alternative process 5 is activated.

  The alternative process management information 72 includes a process ID and a port number for each process number. The process number is identification information that uniquely defines a process that distributes a business request and actually executes a business, in other words, an alternative process 5 that replaces the business process 3 among the activated processes.

  The determination unit 75 may activate the same number of alternative processes 5 as the plurality of business processes 3 when an abnormality occurs in the business process 3. Further, the determination unit 75 may notify the distribution process 2 to distribute the business request to the alternative process 5.

  The business process 3 continues the processing until the execution of the already distributed business is completed. The business process 3 includes the business process 3 in which it is determined that an abnormality has occurred, but has not yet reached an abnormal end. Therefore, the business process 3 can continue the execution of the business already distributed and can be terminated.

  Upon receiving the notification of the completion of the activation of the alternative process 5 and the alternative process management information 72 from the monitoring process 7, the distribution process 2 distributes the business request received from the client 8 to the alternative process 5 based on the alternative process management information 72. Divide. In response to this, the alternative process 5 to which the business request is distributed executes the business requested in the business request by the business application 6.

  As a result, the distribution destination of the business request from the client 8 is changed, and not only the business process 3 in which an abnormality has occurred but all the business processes 3 are switched to the alternative process 5. As a result, not only the business process 3 in which an abnormality has occurred, but also all business processes 3 at the time of the occurrence of the abnormality can be analyzed, so that the cause of the abnormality can be known more accurately. Thereafter, the execution of the business application 6 is continued by the alternative process 5.

  On the other hand, the monitoring process 7 transmits a notification of the completion of the activation of the alternative process 5 to the distribution process 2 and then checks the plurality of business processes 3. The check of the business process 3 is executed by the check unit 76. The check unit 76 checks each of the plurality of business processes 3.

  Specifically, when the alternative process 5 is activated, the check unit 76 performs an abnormality investigation on all the business processes 3 in parallel with the process monitoring process. Specifically, the check unit 76 of the monitoring process 7 requests each of the plurality of business processes 3 to transmit all information. In response to this, the business process 3 transmits all the information about the business process 3 including the business process information at that time to the check unit 76 of the monitoring process 7. Thus, specifically, the monitoring process 7 collects all information about the business process 3 including the business process information from all the business processes 3.

  The check unit 76 may check the plurality of business processes 3 after the determination unit 75 notifies the distribution unit to distribute the business request to the alternative process 5. Further, the check unit 76 may collect information in a predetermined range instead of all the information about the business process 3 including the business process information. When an abnormality occurs in the business process 3, the check unit 76 may perform an abnormality investigation only on the business process 3 in which the abnormality has occurred.

  Thereafter, the administrator of the application server 1 analyzes the collected information. As a result, all business processes # 1 and business processes # 2 are examined in an alive state without affecting other business executions. Can do. Therefore, since the state of the business process 3 when the abnormality has occurred can be maintained, the cause of the abnormality can be correctly clarified.

  FIG. 4 is a diagram illustrating an example of a hardware configuration of the application server 1.

  The CPU 101 controls the application server 1 according to a control program stored in the ROM 102. The CPU 101 executes a program on the RAM 103, which is a main memory, for example. Thereby, the distribution process 2, the business process 3, the business application 4, the alternative process 5, the business application 6, and the monitoring process 7 are realized. For example, the program is stored in a recording medium 109 such as a CD-ROM or DVD, input from the recording medium 109 to the hard disk 106 via a CD-ROM drive or DVD drive, and loaded from the hard disk 106 to the RAM 103.

  The business process management information 71, the alternative process management information 72, and the business process information 31 and 32 are stored in the data storage unit. The data storage unit is provided in the RAM 103 or the hard disk 106, for example. In other words, the data is stored in the RAM 103 or the hard disk 106, for example.

  The input device 104 is a keyboard, for example, and may include a mouse or the like. The output device 105 is a display, for example, and may include an output device such as a printer. The CPU 101, ROM 102, RAM 103, input device 104, output device 105, hard disk 106, and network connection unit 107 are connected to each other via a bus 108.

  The network connection unit 107 is, for example, a transmission / reception device, is connected to the network 9, and is connected to another computer, for example, the client 8 via the network 9. As a result, the application server 1 communicates with the client 8.

  The client 8 is also a computer having the hardware configuration shown in FIG. 4, for example.

  5 and 6 are process monitoring process flows, and integrally represent the process monitoring process flow.

  The monitoring process 7 is started by executing the start command of the monitoring process 7 in the application server 1 (step S1). Furthermore, the distribution process 2 is started by executing a start command for the distribution process 2 in the application server 1 (step S2). The activated distribution process 2 is in a state of waiting for notification of completion of the activation of the business process 3 from the monitoring process 7. Either the activation of the monitoring process 7 or the activation of the distribution process 2 may be executed first.

  The activated monitoring process 7 starts a plurality of business processes 3 by executing a start command for the business processes 3 (step S3). Thereby, for example, the business process # 1 is started (step S4), and the business process # 2 is started (step S5). The activated business process # 1 transmits a notification of its own activation completion to the monitoring process 7, and the activated business process # 2 transmits a notification of its own activation completion to the monitoring process 7.

  When the monitoring process 7 receives notification of activation completion from all the activated business processes # 1 and # 2, it sends a notification of activation completion of the business process 3 to the distribution process 2 (step S6). The business process management information 71 shown in FIG. 3A is added to the notification of the start completion of the business process 3 to the distribution process 2 in step S6.

  When the distribution process 2 receives the notification of the completion of the activation of the business process 3 from the monitoring process 7, it starts accepting the business request from the client 8, and the received business request is equally distributed to the business process # 1 and the business process # 2. Distribution is performed (step S7). Thereafter, the business application 4 is executed by the business process # 1 and the business process # 2.

  The business process # 1 to which the business request has been distributed executes the business requested in the business request by the business application (step S8). Further, the business process # 2 to which the business request is distributed executes the business requested in the business request by the business application (step S9).

  On the other hand, the collection unit 73 of the monitoring process 7 transmits a notice of the completion of the activation of the business process 3 to the distribution process 2 in step S6, and then periodically, all the business processes # 1 and business processes are periodically performed at a predetermined cycle. Business process information is collected from # 2 (step S10). For example, the business process information 31 shown in FIG. 3B is collected from the business process # 1, and the business process information 32 shown in FIG. 3C is collected from the business process # 2.

  Next, when the comparison unit 74 of the monitoring process 7 collects the business process information from all the business processes # 1 and # 2, it compares the collected business process information (step S11).

  Next, as a result of the comparison in step S <b> 11, the determination unit 75 of the monitoring process 7 determines that the first business process 3 characteristic information is different from the characteristic indicated by the characteristic information set of the second business process 3. It is determined that an abnormality has occurred in the business process 3 (step S12).

  Next, when the monitoring process 7 determines that an abnormality has occurred in the business process 3, for example, it alerts the administrator of the application server 1 that an abnormality has occurred in the business process 3 (step S13).

  Next, the monitoring process 7 activates the same number of alternative processes 5 as the number of business processes 3 (step S14). Thereby, for example, the alternative process # 1 corresponding to the business process # 1 is started (step S15), and the alternative process # 2 corresponding to the business process # 2 is started (step S16). In other words, when it is determined that an abnormality has occurred in the business process 3, not only the business process 3 in which the abnormality has occurred but all the business processes 3 are switched to the alternative process 5 as described later. The activated alternative process # 1 transmits its own activation completion notification to the monitoring process 7, and the activated alternative process # 2 transmits its own activation completion notification to the monitoring process 7.

  The activation of the alternative process 5 may be executed at various timings. Before the occurrence of abnormality, for example, the alternative process 5 may be started in step S3. Alternatively, step S14 may be executed after step S12, and then step S13 may be executed.

  In addition, when it is determined that an abnormality has occurred in the business process 3, only the business process 3 in which the abnormality has occurred may be switched to the alternative process 5.

  When the monitoring process 7 receives notification of activation completion from all the activated alternative processes # 1 and # 2, it sends a notification of change of the distribution destination to the distribution process 2 (step S17). The alternative process management information 72 shown in FIG. 3D is added to the notification of the change of the distribution destination to the distribution process 2 in step S17, in other words, the notification of the start completion of the alternative process 5.

  Upon receiving the notification of the completion of the activation of the alternative process 5 from the monitoring process 7, the distribution process 2 changes the distribution destination of the business request from the client 8 from the business process 3 to the alternative process 5, and receives the business request received from the client 8. Then, the process is equally distributed to the alternative process # 1 and the alternative process # 2 (step S18). As a result, all the business processes 3 are switched to the alternative process 5. Thereafter, the execution of the business application 6 is continued by the alternative process # 1 and the alternative process # 2.

  FIG. 7 is a process investigation process flow.

  When the alternative process 5 is activated, the check unit 76 of the monitoring process 7 performs an abnormality investigation on all the business processes 3 in parallel with the process monitoring process (step S21). As a result, the abnormality is also investigated for the business process # 1 and the business process # 2 that cause the activation of the alternative process # 1 and the alternative process # 2. Specifically, the check unit 76 of the monitoring process 7 requests the business process # 1 to transmit all the information about the business process # 1, and sends the business process # 2 all about the business process # 2. Request transmission of information.

  In response to this, the business process # 1 transmits all information about the business process # 1 including the current business process information to the check unit 76 of the monitoring process 7 (step S22). Also, the business process # 2 transmits all information regarding the business process # 2 including the current business process information to the check unit 76 of the monitoring process 7 (step S23).

  Thereafter, the administrator of the application server 1 analyzes the collected information, and in the state where all the business processes # 1 and # 2 are alive, all the business processes # 1 and # 2 are processed. Can be investigated.

DESCRIPTION OF SYMBOLS 1 Application server 2 Distribution process 3 Business process 4, 6 Business application 5 Alternative process 7 Monitoring process 8 Client 9 Network 73 Collection part 74 Comparison part 75 Judgment part 76 Check part

Claims (8)

For each of a plurality of processes that execute processing based on a plurality of application programs, a collection unit that collects characteristic information representing the characteristics of the process;
An abnormality occurs in a certain process when the characteristic information related to a certain process of the plurality of processes is different from the characteristic information of any of the plurality of other processes different from the certain process. An information processing apparatus comprising: a determination unit that determines that the determination has been made. The information processing apparatus further includes:
A distribution unit that distributes requests from clients to the plurality of processes;
When the determination unit determines that an abnormality has occurred in the one or more processes, the distribution unit starts the same number of alternative processes as the plurality of processes and distributes the request to the alternative processes. The information processing apparatus according to claim 1, wherein the information processing apparatus is notified. The information processing apparatus further includes:
The information processing apparatus according to claim 2 , further comprising: a check unit that checks the plurality of processes after the determination unit notifies the distribution unit to distribute the request to the alternative process. The characteristic information is information indicating the number of times a result of processing executed by a process is a predetermined processing result or information on a usage amount of a predetermined resource used by the process. Information processing device. The information processing apparatus according to claim 4, wherein the characteristic information is the number of times a status code other than the status code 200 in HTTP, which is a protocol when the plurality of processes communicate, is generated. The monitoring process collects characteristic information that represents the characteristics of the process for each of a plurality of processes that execute processing based on a plurality of application programs.
The monitoring process includes the case where the characteristic information related to a certain process of the plurality of processes is different from the characteristic information of any of a plurality of other processes different from the certain process among the plurality of processes. A process monitoring method characterized by determining that an abnormality has occurred in a process. A process monitoring program for monitoring a process,
The program is stored in a computer.
For each of a plurality of processes that execute processing based on a plurality of application programs, processing for collecting characteristic information representing the characteristics of the process;
An abnormality occurs in a certain process when the characteristic information related to a certain process of the plurality of processes is different from the characteristic information of any of the plurality of other processes different from the certain process. A process monitoring program characterized by causing a process to be judged to have been executed. A recording medium for recording a process monitoring program for monitoring a process,
The process monitoring program is stored in a computer.
For each of a plurality of processes that execute processing based on a plurality of application programs, processing for collecting characteristic information representing the characteristics of the process;
An abnormality occurs in a certain process when the characteristic information related to a certain process of the plurality of processes is different from the characteristic information of any of the plurality of other processes different from the certain process. A recording medium characterized by causing a process to be determined to have been performed.

Images