JP2012138015A - Time-out monitoring system and time-out monitoring program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a time-out monitoring system and a time-out monitoring program for early detection of time-out at the time of request processing execution with plural processes working together.SOLUTION: A timer monitoring unit 10 of a process device A specifies a transaction request and specifies a timer value according to the transaction request. The process device A transmits a telegraphic message including the timer value to each succeeding process device B to D, and the timer monitoring unit 10 executes event monitoring processing. The timer monitoring unit 10 determines whether a process is normally complete or time-out occurs, and when determining that time-out occurs, executes time-out notification processing. Timer monitoring units 20 to 40 of the process devices B to D each obtain a timer value from a preceding process. The timer monitoring units 20 to 40 each calculate a new timer value obtained by subtracting a subtraction value from the timer value and execute telegraphic message transmission processing including the timer value.

Description

  The present invention relates to a timeout monitoring system and a timeout monitoring program for managing the progress of a process and detecting a timeout of each process.

  In a computer system, a plurality of processes may be linked to perform one business process (see, for example, Patent Document 1). In the technique described in this document, data from a client terminal is transferred to a host system. In this case, a system composed of a plurality of communication relay devices and a business management server is used. In such a system environment, if a specific request process takes time, or if the process cannot proceed due to a failure, it is not possible to secure resources for processing other requests. . Therefore, there is often a time limit for completing each request process. When this time limit is exceeded, it is determined as a timeout, resources are released, and the amount of resources used for other request processing performed in parallel is secured.

  For this reason, in order to quickly recover from a failure even when a large number of processing requests stay, a method for setting a processing termination time limit has been studied (for example, see Patent Document 2). In the technique described in this document, the response monitoring timeout time is added to the processing request data to the call destination when RPC is issued from the call source. Then, the response monitoring timeout time is taken over between the servers. In the middle and subsequent servers, when further processing is requested to another server, the difference between the caller timeout time and the current time is set as the response monitoring time. Also, if the call source timeout time has passed or if there is no response from the call destination even after the response monitoring time has passed, the processing is terminated.

JP 2009-237789 A (first page, FIG. 2) JP 2000-40020 (first page, FIG. 5)

  For such processing, it is desirable that a timeout can be detected as early as possible. On the other hand, there is a case where a plurality of processes are linked and processed for one transaction processing. In this case, it is necessary to consider the time required for processing each process. Therefore, as in Patent Document 2, if only the remaining time with respect to the response monitoring timeout time is monitored, the timeout cannot be detected until this response monitoring timeout time is reached, and it is not possible to respond quickly.

  Furthermore, a process (cooperation process) necessary for processing each request differs depending on the content of the request. Therefore, there is a mismatch that the total time of the time limit set in each process becomes longer than the time limit set in the first processing device when the time limit is individually set according to each process. It may occur. In particular, when the number of linkages is large, the total time becomes long, the purpose of early response cannot be achieved, and inconsistency is likely to occur with the default time limit.

In the cooperation process, a timeout time is set in consideration of a predicted time assuming a required time for each process. When the number of linkages is large, there is a high possibility that the gap between the actual required time and the predicted time will increase during the linkage. Therefore, it is desirable that a more accurate remaining time can be calculated during the cooperation process.

  When processing a request for a transaction (service), various channels such as store sales and online sales may be used. Thus, it is desirable to manage timeouts at the same time for common transactions even when the channels are different.

  The present invention has been made in view of the above-described problems, and an object of the present invention is to provide a timeout monitoring system and a timeout monitoring program for early detection of timeout when request processing is executed in cooperation with a plurality of processes. Is to provide.

  In order to solve the above problem, the invention according to claim 1 executes a takeover process of transmitting a second message for requesting the subsequent process to be processed in accordance with the first message received from the preceding process. And a process execution unit that receives a return message from the subsequent process, and a timer monitoring unit that is connected to a timer that notifies the elapse of a timer value and acquires event information from the process execution unit and the timer. In the timeout monitoring system, the timer monitoring unit obtains a first timer value from the first message received by the process execution unit, and subtracts an assumed time necessary for the takeover process from the first timer value. 2 timer value setting means for calculating the timer value and setting the second timer value in a second message to be taken over by the process execution unit to the succeeding process; Monitoring means for waiting for the elapse of the second timer value and the return message from the subsequent process in the process execution unit, and detecting the timeout of the second timer value before receiving the return message Includes a time-out report unit that outputs a time-out report and allows the process execution unit to perform processing based on the return message when the return message is received before the second timer value times out. This is the gist.

  The invention according to claim 2 is summarized in that, in the timeout monitoring system according to claim 1, the timer monitoring unit holds a predetermined value in advance as an estimated time required for the takeover process.

  According to a third aspect of the present invention, in the time-out monitoring system according to the first or second aspect, the timer monitoring unit determines whether or not a subsequent linked process is necessary based on a return message received by the process execution unit. When a new linkage process is performed, a remaining time obtained by subtracting an elapsed time until the return message is received from the first timer value is calculated as a second timer value. The gist is to set the second timer value in the second message taken over by the linkage process.

  According to a fourth aspect of the present invention, in the time-out monitoring system according to the third aspect, the timer monitoring unit, when there is a subsequent cooperative process, a remaining time and a reference for performing the subsequent cooperative process When the remaining time is equal to or less than a reference value, a time-out report is output.

  According to a fifth aspect of the present invention, in the timeout monitoring system according to any one of the first to fourth aspects, when the timer monitoring unit first receives a transaction request, the timer monitoring unit is based on the content of the transaction request. The gist is to set the initial value of the timer value.

According to a sixth aspect of the present invention, in response to the first message received from the preceding process, the handover process for transmitting the second message for requesting the subsequent process is executed, and the return message from the subsequent process is executed. Using a time-out monitoring system that is connected to a process execution unit that receives the timer value and a timer that notifies the elapse of the timer value and includes event monitoring information from the process execution unit and the timer. A program for monitoring, wherein the timer monitoring unit acquires a first timer value from a first message received by the process execution unit, and subtracts an estimated time required for the takeover process from the first timer value A timer value for setting the second timer value in a second message that is calculated by the process execution unit and is taken over by the subsequent process in the process execution unit Monitoring means for waiting for the elapse of the second timer value in the timer and the return message from the succeeding process in the process execution unit; timeout of the second timer value before receiving the return message A time-out report that outputs a time-out report when the second timer value is timed out, and that accepts processing based on the return message in the process execution unit when the return message is received before the time-out of the second timer value. It is made to function as.

(Function)
According to the first or sixth aspect of the invention, the timer monitoring unit acquires the first timer value from the first message received by the process execution unit, and calculates the estimated time required for the takeover process from the first timer value. The subtracted second timer value is calculated, and the second timer value is set in the second telegram to be taken over by the subsequent process in the process execution unit. And it waits for progress of the said 2nd timer value in a timer, and the return message from the subsequent process in a process execution part. If a timeout of the second timer value is detected before receiving a return message, a timeout report is output. On the other hand, if a return message is received before the second timer value times out, the process execution unit allows processing based on the return message. Accordingly, it is possible to take over a timer value for detecting a timeout in a subsequent process in consideration of an assumed time required for the takeover process. Therefore, even when a plurality of processes are linked, a timeout can be detected quickly.

  According to the second aspect of the present invention, the timer monitoring unit can set a timer value used in the subsequent process by holding a predetermined value in advance as the estimated time required for the takeover process.

  According to the third aspect of the present invention, the timer monitoring unit determines whether or not a subsequent cooperative process is necessary based on the return message received by the process execution unit. When a new linkage process is performed, the remaining time obtained by subtracting the elapsed time until the return message is received from the first timer value is calculated as the second timer value, and the process execution unit executes the new linkage process. The second timer value is set in the second message to be taken over. Thereby, the timeout can be detected quickly by setting the timer value according to the necessity of the subsequent cooperation process.

  According to the fourth aspect of the present invention, when the subsequent cooperative process remains, the timer monitoring unit compares the remaining time for performing the subsequent cooperative process with the reference value. If the remaining time is less than the reference value, a timeout report is output. Thereby, when there are many remaining cooperation processes with respect to the remaining time until timeout, timeout can be estimated and it can respond quickly.

  According to the fifth aspect of the present invention, the timer monitoring unit sets the initial value of the timer value based on the content of the transaction request when the transaction request is first received. Thereby, an appropriate initial value of the timer value can be set according to the transaction content.

According to the present invention, it is possible to provide a timeout monitoring system and a timeout monitoring program for detecting a timeout at an early stage when request processing is executed in cooperation with a plurality of processes.

The system schematic of embodiment of this invention. Explanatory drawing of the functional block of the timer monitoring part of embodiment of this invention. It is explanatory drawing of the processing procedure of this embodiment, Comprising: (a) is a processing procedure in the process apparatus A, (b) is explanatory drawing of the processing procedure in the process apparatus (B, D). Explanatory drawing of the process sequence in the process apparatus C of this embodiment. It is explanatory drawing of the cooperation process of this embodiment, Comprising: (a) is explanatory drawing of the message | telegram transmitted / received between process apparatuses, (b) is explanatory drawing of the timer value in each process apparatus. Explanatory drawing of the process sequence of other embodiment. Explanatory drawing of the process sequence of other embodiment.

  Hereinafter, an embodiment embodying the present invention will be described with reference to FIGS. In the present embodiment, it is assumed that a transaction request is processed by a plurality of processes in the process devices A to E as shown in FIG. The process devices A to E include process execution units A1 to E1. Here, based on the received message (first message), the process execution units A1 to D1 transmit a message (second message) requesting processing to the process device that executes the subsequent process. Then, the process execution units B1 to E1 return a return message to the preceding process execution units A1 to D1.

  The process device A is a computer process device that identifies a transaction request message input by a person in charge or a customer. The process execution unit A1 of the process device A acquires a transaction request message from an automatic teller machine (ATM) installed at each base of the bank, a front terminal installed at a store, or the like. Then, the process execution unit A1 assigns a transaction type identifier to the transaction request message, and associates the customer identifier and transaction content included in the transaction request message with the transaction type identifier in a transaction information storage unit (not shown). Record. Then, the process execution unit A1 transfers this transaction request message to the process device B. Further, when receiving a transaction response message corresponding to the transaction request message from the process device B, the process execution unit A1 outputs the transaction request message to the terminal that transmitted the transaction request message.

  The process device B is a computer device (hub device) that relays a message between the process device A and the process device C. The process execution unit B1 of the process device B transfers the transaction request message received from the process device A to the process device C. In addition, the process execution unit B1 transfers the transaction response message received from the process device C to the process device A.

  The process device C is a computer device that identifies a necessary cooperation process (cooperation processing) based on the contents of a transaction request message and transmits a processing request message to a business component that executes this process. The process execution unit C1 of the process apparatus C holds a transmission destination table. Then, using this transmission destination table, the business component of the transmission destination and the content of the processing request are determined based on the business logic in accordance with the content of the received transaction request message. The process execution unit C1 transmits a processing request message to the process device E including the business component via the process device D, and receives a processing result message from the process device E.

Further, when the subsequent cooperation process (cooperation process) is necessary according to the content of the transaction request message, the process execution unit C1 newly generates a process request message and transmits the process request message to the business component. Here, the cooperative processing is processing in which a plurality of individual transactions (for example, balance check, account existence check, payment processing, etc.) are continuously handled as a series of transactions. And when all the cooperation processes corresponding to a transaction request message are complete | finished, the transaction response message corresponding to a transaction request message is returned to the process apparatus A via the process apparatus B.

  The process device D is a computer device (hub device) that relays a message between the process device C and the process device E. The process execution unit D1 of the process device D transfers the processing request message received from the process device C to the process device E. In addition, the process execution unit D1 transfers the processing result message received from the process device E to the process device C.

  The process device E is a computer device (for example, a host computer) that executes each process necessary for handling a processing request message. For example, the process execution unit E1 of the process device E includes business components for executing various processes such as an account existence check process, a balance check process, a payment process, a deposit process, and an account accounting process.

  Further, the process devices A to D include a timer TM and timer monitoring units 10 to 40 as shown in FIG. The timer monitoring units 10 to 40 monitor timeouts in the process devices A to D.

  The timer monitoring unit 10 executes processing for monitoring a timeout for a transaction request. The timer monitoring unit 10 includes control means (CPU, RAM, ROM, and the like), and performs processing described later (processing including a timer value setting stage, an event monitoring stage, and a timeout reporting stage). Therefore, by executing the process management program, the timer monitoring unit 10 functions as the timer value setting unit 111, the event monitoring unit 112, and the timeout reporting unit 113.

  The timer value setting means 111 executes processing for setting the remaining time (timer value) until timeout in accordance with the transaction request message. The timer value setting means 111 holds a timer initial value table. In this timer initial value table, initial values of timer values are stored in association with customer identifiers and transaction type identifiers. The customer identifier and the transaction type identifier are identifiers for specifying the customer and the transaction type of the trading partner, respectively. The timer initial value is a time limit for completing the transaction request.

  The event monitoring unit 112 executes a progress monitoring process of the timer TM or the process execution unit A1. The event monitoring unit 112 acquires event information about return message reception at the process execution unit A1 or time-up at the timer TM.

  The time-out reporting unit 113 executes processing for notifying the process execution unit A1 that is a transmission source specified by the transaction request message when time-out occurs. Then, based on this timeout notification, a transaction cancellation message for a transaction request is transmitted to each process device, and the resources used to respond to the transaction request are released.

  The timer monitoring unit (20, 40) executes processing for monitoring timeout based on the received message. The timer monitoring units (20, 40) are provided with control means (CPU, RAM, ROM, etc.) and perform processing (processing including a timer value setting stage, an event monitoring stage, and a timeout reporting stage) described later. Therefore, by executing the process management program, the timer monitoring units (20, 40) serve as timer value setting means (211 and 411), event monitoring means (212 and 412), and timeout report means (213 and 413). Function.

The timer value setting means (211 and 411) executes processing for setting the remaining time (timer value) until timeout in accordance with the transaction request message and the processing request message. The timer value setting unit 211 holds a subtraction value table. This subtraction value table stores a subtraction value for shortening the timer value. The subtraction value is an estimated time (predetermined time) required for the handover process for the process execution unit (B1, D1) of the process device (B, D) as the preceding process to request the subsequent process.

Similar to the event monitoring unit 112, the event monitoring unit (212, 412) executes a progress monitoring process of the timer TM or the process execution unit (B1, D1).
Similar to the timeout report unit 113, the timeout report unit (213, 413) executes processing for notifying a timeout when a timeout occurs.

  The timer monitoring unit 30 executes processing for monitoring a timeout for a received message. The timer monitoring unit 30 includes control means (CPU, RAM, ROM, and the like), and performs processing described later (processing including a timer value setting stage, an event monitoring stage, and a timeout reporting stage). Therefore, by executing the process management program, the timer monitoring unit 30 functions as the timer value setting unit 311, the event monitoring unit 312, and the timeout reporting unit 313.

  The timer value setting unit 311 executes processing for setting the remaining time (timer value) until timeout in accordance with the processing request message. The timer value setting means 311 holds a subtraction value table, similarly to the timer value setting means (211 and 411). The timer value setting unit 311 includes a temporary storage memory for recording the acquired timer value. Then, when the subsequent cooperation processing is necessary, the timer value setting unit 311 obtains a new timer value by subtracting the time required to receive the return message from the timer value recorded in the temporary storage memory. Set as.

Similar to the event monitoring unit 112, the event monitoring unit 312 executes a monitoring process of the progress status of the timer TM or the process execution unit C1.
Similar to the timeout report unit 113, the timeout report unit 313 executes a process of notifying a timeout when a timeout occurs.

  Next, a procedure for process management in the system configured as described above will be described with reference to FIGS. Here, as shown in FIG. 5A, the transaction request of the process apparatus A is transferred to the process apparatus C via the process apparatus B. Then, it is assumed that the process device C processes the process device E via the process device D in three linkages and returns a transaction response message to the process device A via the process device B. In this case, timer monitoring processing is performed in the process devices A to D.

(Timer monitoring process in process device A)
First, the timer monitoring process in the process apparatus A will be described with reference to FIG.
Here, the timer monitoring unit 10 of the process device A executes transaction request specifying processing (step S1-1). Specifically, the timer value setting unit 111 of the timer monitoring unit 10 acquires information on the transaction type and customer identifier for the transaction request acquired in the process execution unit A1.

  Next, the timer monitoring unit 10 of the process device A executes a timer value specifying process in response to the transaction request (step S1-2). Specifically, the timer value setting unit 111 of the timer monitoring unit 10 extracts a timer value associated with the acquired transaction type and customer identifier from the timer initial value table. In the present embodiment, as shown in FIG. 5B, “118 seconds” is used as the initial timer value.

Next, the process device A executes a message transmission process including a timer value (step S1-3). Specifically, the process execution unit A1 of the process device A creates a transaction request message to be transmitted to the process device B. In this case, the timer value setting unit 211 of the timer monitoring unit 20 stores the extracted timer value “118 seconds” in the timer value area of the transaction request message transmitted to the process device B. Then, this transaction request message is transmitted to the process device B by the process execution unit A1. The timer value setting unit 111 supplies the timer value “118 seconds” to the timer TM of the process apparatus A. In this case, the timer TM starts counting down from the timer value “118 seconds”.

  Then, the timer monitoring unit 10 of the process device A executes event monitoring processing (step S1-4). Specifically, the timer TM supplies a time-up notification to the event monitoring unit 112 of the timer monitoring unit 10 when the timer value “118 seconds” has been counted down. When the process execution unit A1 receives a return message (transaction response message) from the process device B, the process execution unit A1 supplies a return message reception notification to the event monitoring unit 112 of the timer monitoring unit 10. In order to acquire such event information, the event monitoring unit 112 of the timer monitoring unit 10 waits for a time-up notification or a return message reception notification in the timer TM.

  Then, the timer monitoring unit 10 of the process apparatus A executes normal completion or timeout determination processing (step S1-5). Specifically, the event monitoring unit 112 of the timer monitoring unit 10 determines the timeout based on the relationship between the time-up notification from the timer TM or the return message reception notification from the process execution unit A1. Here, if a time-up notification is acquired before the return message reception notification, it is determined that a timeout has occurred. On the other hand, when a return message notification is acquired before the time-up notification, it is determined that the operation is normally completed.

If it is determined that a timeout has occurred ("timeout" in step S1-5), the timer monitoring unit 10 of the process device A executes a timeout report process (step S1-6). Specifically, the event monitoring unit 112 of the timer monitoring unit 10 notifies the timeout reporting unit 113 of the timeout. In this case, the timeout report means 113 notifies the timeout to the sender of the transaction request message.
On the other hand, if it is determined as normal completion (in the case of “normal completion” in step S1-5), the timer monitoring unit 10 of the process apparatus A ends the timer monitoring processing.

(Timer monitoring process in process device B)
Next, the timer monitoring process in the process apparatus B will be described with reference to FIG. In this process device B, the transaction request message received from the process device A is transmitted to the process device C, and the timer is monitored until the transaction response message is received from the process device C.

  The timer monitoring unit 20 of the process device B executes a timer value acquisition process from the preceding process (step S2-1). Specifically, the timer value setting unit 211 of the timer monitoring unit 20 acquires a timer value (first timer value) from the timer value area of the transaction request message acquired from the preceding process device A. In the present embodiment, “118 seconds” is acquired as the timer value, as shown in FIG.

  Next, the timer monitoring unit 20 of the process device B executes a subtraction process of the subtraction value from the timer value (step S2-2). Specifically, the timer value setting unit 211 of the timer monitoring unit 20 subtracts the subtraction value stored in the subtraction value table from the acquired timer value. In the present embodiment, “3 seconds” is acquired as the subtraction value. Then, the timer value setting unit 211 calculates “115 seconds” obtained by subtracting “3 seconds” from “118 seconds” as a new timer value (second timer value).

Next, the process device B executes a message transmission process including a timer value (step S2-3). Specifically, the process execution unit B1 of the process device B prepares a transaction request message to be transferred to the process device C. In this case, the timer value setting unit 211 of the timer monitoring unit 20 stores the calculated timer value “115 seconds” in the timer value area of the transaction request message transferred to the process device C. This transaction request message is transferred to the subsequent process device C by the process execution unit B1. Further, the timer value setting means 211 supplies the timer value “115 seconds” to the timer TM of the process apparatus B. In this case, the timer TM starts counting down from the timer value “115 seconds”.

  Then, the timer monitoring unit 20 of the process device B executes event monitoring processing (step S2-4). Specifically, when the timer TM has finished counting down the timer value “115 seconds”, the timer TM supplies a time-up notification to the event monitoring unit 212 of the timer monitoring unit 20. When the process execution unit B1 receives a return message (transaction response message) from the process device C, the process execution unit B1 supplies a return message reception notification to the event monitoring unit 212 of the timer monitoring unit 20. In order to acquire such event information, the event monitoring unit 212 of the timer monitoring unit 20 waits for a time-up notification or a return message reception notification in the timer TM.

  Then, the timer monitoring unit 20 of the process device B executes normal completion or timeout determination processing (step S2-5). Specifically, the event monitoring unit 212 of the timer monitoring unit 20 determines the timeout based on the prior relationship of the time-up notification from the timer TM or the return message reception notification from the process execution unit B1. Here, if a time-up notification is acquired before the return message reception notification, it is determined that a timeout has occurred. On the other hand, when a return message notification is acquired before the time-up notification, it is determined that the operation is normally completed.

If it is determined that a timeout has occurred ("timeout" in step S2-5), the timer monitoring unit 20 of the process apparatus B executes a timeout report process (step S2-6). Specifically, the event monitoring unit 212 of the timer monitoring unit 20 notifies the timeout reporting unit 213 of the timeout. In this case, the timeout report means 213 notifies the transaction request transmission source of the timeout.
On the other hand, when it is determined that the operation is normally completed (in the case of “normal completion” in step S2-5), the timer monitoring unit 20 of the process apparatus B ends the timer monitoring process.

  In the process apparatus D, the same timer monitoring process as that of the process apparatus B is performed. In this process apparatus D, a timer is monitored until a process result message for the process request message received from the process apparatus C is received from the process apparatus E. As shown in FIG. 5B, since “2 seconds” is set as the subtraction value in the process apparatus C, the timer value obtained by subtracting “2 seconds” from the timer value acquired from the process apparatus C Monitor.

(Timer monitoring process in process device C)
Next, timer monitoring processing in the process apparatus C will be described with reference to FIG.
The timer monitoring unit 30 of the process apparatus C executes timer value acquisition processing from the preceding process (step S3-1). Specifically, the timer value setting unit 311 of the timer monitoring unit 30 acquires a timer value (first timer value) from the timer value area of the transaction request message acquired from the preceding process device B. In this embodiment, as shown in FIG. 5B, “115 seconds” is acquired as the timer value. In this case, the timer value setting unit 311 records the acquired timer value in the temporary storage memory. As will be described later, the timer value recorded in the temporary storage memory is used when calculating the timer value for the second and subsequent linkage processes.

Next, the timer monitoring unit 30 of the process device C executes a subtraction process of the subtraction value from the timer value (step S3-2). Specifically, the timer value setting unit 311 of the timer monitoring unit 30 subtracts the subtraction value stored in the subtraction value table from the acquired timer value. In this embodiment, “2 seconds” is acquired as the subtraction value. Then, the timer value setting unit 211 calculates “113 seconds” obtained by subtracting “2 seconds” from “115 seconds” as a new timer value (second timer value).

  Next, the process device C executes a message transmission process including a timer value (step S3-3). Specifically, the process execution unit C1 of the process device C generates a processing request message to be transmitted to the subsequent process using the transmission destination table according to the content of the transaction request message. Then, the timer value setting unit 311 of the timer monitoring unit 30 stores the calculated timer value “113 seconds” in the timer value area of the processing request message to be transmitted to the process device D. This processing request message is transmitted to the subsequent process device D by the process execution unit C1. The timer value setting means 311 supplies the timer value “113 seconds” to the timer TM. In this case, the timer TM starts counting down from the timer value “113 seconds”.

  Then, the timer monitoring unit 30 of the process device C executes event monitoring processing (step S3-4). Specifically, the timer TM supplies a time-up notification to the event monitoring unit 312 of the timer monitoring unit 30 when the count-down of the timer value “113 seconds” is completed. When the process execution unit C1 receives a return message (processing result message) from the process device D, the process execution unit C1 supplies a return message reception notification to the event monitoring unit 312 of the timer monitoring unit 30. In order to acquire such event information, the event monitoring unit 312 of the timer monitoring unit 30 waits for a time-up notification or a return message reception notification in the timer TM.

  Then, the timer monitoring unit 30 of the process apparatus C executes normal completion or timeout determination processing (step S3-5). Specifically, the event monitoring unit 312 of the timer monitoring unit 30 determines the timeout based on the prior relationship of the time-up notification from the timer TM or the return message reception notification from the process execution unit C1. Here, if a time-up notification is acquired before the return message reception notification, it is determined that a timeout has occurred. On the other hand, when a return message notification is acquired before the time-up notification, it is determined that the operation is normally completed.

  If it is determined that a timeout has occurred (in the case of “timeout” in step S3-5), the timer monitoring unit 30 of the process device C executes a timeout report process (step S3-6). Specifically, the event monitoring unit 312 of the timer monitoring unit 30 notifies the timeout reporting unit 313 of the timeout. In this case, the timeout report means 313 notifies the transaction request transmission source of the timeout.

  On the other hand, when it is determined that the process is normally completed (in the case of “normally completed” in step S3-5), the timer monitoring unit 30 of the process device C executes a determination process as to whether or not a subsequent cooperation process is necessary ( Step S3-7). Specifically, the process execution unit C1 of the process apparatus C uses the transmission destination table to determine whether or not subsequent cooperative processing is necessary based on the received processing result message. Then, the timer value setting unit 311 of the timer monitoring unit 30 acquires information on necessity of new cooperation processing from the process execution unit C1.

When it is determined that the subsequent cooperation processing is necessary based on the necessity information of the cooperation processing acquired from the process execution unit C1 (in the case of “YES” in step S3-7), the timer monitoring unit 30 of the process device C A new timer value calculation process is executed (step S3-8). Specifically, the timer value setting means 311 of the timer monitoring unit 30 takes the time required for the processing result message to return from the timer value at the time of transmission of the processing request message (the timer value recorded in the temporary storage memory). A new timer value is calculated by subtracting. In this case, the timer value setting unit 311 updates the timer value in the temporary storage memory using the acquired timer value. In this embodiment, in FIG.5 (b), the case where a return message | telegram is received 3 seconds after the first cooperation process is assumed. In this case, a value “110 seconds” obtained by subtracting the required time “3 seconds” from the timer value “113 seconds” first acquired by the process apparatus C is used as the timer value acquired from the preceding process (cooperation processing). Use. In the second linkage process, it is assumed that a return message is received after 10 seconds. In this case, a value “100 seconds” obtained by subtracting the required time “10 seconds” from the timer value “110 seconds” of the preceding process (cooperation processing) is used. And the timer monitoring part 30 of the process apparatus C repeats the process after step S3-3 using the calculated new timer value.

  In FIG. 5B, in the first to third linked processing, the process device D receives the timer values “113 seconds”, “110 seconds”, “100 seconds” (first timer value) from the process device C. Therefore, the timer monitoring is performed using “111 seconds”, “108 seconds”, and “98 seconds” (second timer value) obtained by subtracting the subtraction value “2 seconds”.

  On the other hand, when it is determined that there is no subsequent cooperation process based on the necessity information of the cooperation process acquired from the process execution unit C1 (in the case of “NO” in step S3-7), the timer monitoring unit 30 of the process apparatus C Ends the timer monitoring process.

As described above, according to the present embodiment, the following effects can be obtained.
(1) According to the present embodiment, processing consisting of a plurality of processes is performed in the process devices A to E for transaction requests. The process devices A to D are provided with a timer TM and timer monitoring units 10 to 40, and each process device A to D monitors a timeout. Thereby, in each process performed continuously, it is possible to quickly detect a timeout. Then, it is possible to promptly respond to this timeout (such as resource release).

  (2) According to the present embodiment, the timer value is taken over from the preceding process device. As a result, the timer value based on the initially set initial timer value is successively taken over, so that a timeout can be detected in each process in consideration of the time limit for one transaction request. The timer value setting means 111 holds a timer initial value table. In this timer initial value table, initial values of timer values are stored in association with customer identifiers and transaction type identifiers. Depending on the customer and transaction type, the number of linkages for responding to a transaction request may be different. Also, depending on the transaction, it may be better to extend the time limit and complete the transaction request. By using the timer initial value table, it is possible to set a time limit according to the customer and transaction type.

  (3) According to the present embodiment, the timer monitoring units 20 to 40 of the process devices B to D execute the subtraction process of the subtraction value from the timer value (Steps S2-2 and S3-2). Thus, the timer value can be taken over to the next process in consideration of the time required for each process.

  (4) According to the present embodiment, when it is determined that the subsequent cooperation process is necessary based on the necessity information of the cooperation process acquired from the process execution unit C1 (in the case of “YES” in step S3-7), The timer monitoring unit 30 of the process apparatus C executes a new timer value calculation process (step S3-8). Thereby, when a new cooperation process is required, the timer value can be taken over to the next process in consideration of the remaining time when the return message is received.

In addition, you may change the said embodiment as follows.
In the above embodiment, when it is determined that the subsequent cooperation process is necessary based on the necessity information of the cooperation process acquired from the process execution unit C1 (in the case of “YES” in step S3-7), the process device C The timer monitoring unit 30 executes a new timer value calculation process (step S3-8). Here, the timeout may be determined according to the remaining number of linkages. This process will be described with reference to FIG. In this case, the timer value setting means 311 stores a required time (assumed required time) that is assumed to be necessary in one linked process.

Here, similarly to FIG. 4, the timer value acquisition process (step S3-1) from the preceding process to the determination process (step S3-7) as to whether or not the subsequent linkage process is necessary.
Then, when it is determined that the subsequent cooperative processing is necessary based on the necessity information of the cooperative processing acquired from the process execution unit C1 (in the case of “YES” in step S3-7), the timer monitoring unit 30 of the process device C Performs the process of specifying the remaining number of linkages (step S4-1). Specifically, the timer value setting unit 311 of the timer monitoring unit 30 acquires the number of unprocessed linkages that are necessary for completing the transaction request from the process execution unit C1.

  Next, the timer monitoring unit 30 of the process device C executes a determination process as to whether or not the remaining timer value is sufficient for the remaining number of cooperation (step S4-2). Specifically, the timer value setting unit 311 of the timer monitoring unit 30 calculates the required time by multiplying the estimated required time by the remaining number of linkages. Next, the timer value setting unit 311 compares the current timer value with the calculated required time. Here, if the current timer value is shorter than the required time, it is determined that the timer value is not sufficient.

  When it is determined that the remaining timer value is not sufficient for the remaining number of linkages (in the case of “NO” in step S4-2), the timer monitoring unit 30 of the process device C executes a timeout report process (step S3- 6).

  On the other hand, when it is determined that the remaining timer value is sufficient with respect to the remaining number of linkages (in the case of “YES” in step S4-2), the timer monitoring unit 30 of the process apparatus C executes a new timer value calculation process (Step S3-8).

  As a result, in a situation where the timer value becomes small and the remaining time is small, it is possible to determine that a time-out has occurred when many linkage processes remain. Accordingly, it is possible to predict the occurrence of timeout at an early stage and respond quickly.

  Note that the determination method as to whether the remaining timer value is sufficient for the remaining number of linkages is not limited to the method using the estimated required time. For example, the timer value setting unit 311 may hold a timeout determination table in which timer reference values for determining timeout are stored for the number of linkages. In this case, when the timer value setting unit 311 acquires the remaining number of linkages, the timer reference value is acquired using the timeout determination table and compared with the current timer value to determine the timeout.

  In the above embodiment, the timer value setting unit 111 holds a timer initial value table in which initial values of timer values are stored. Further, the timer value setting means (211, 311, 411) holds a subtraction value table in which a subtraction value for shortening the timer value is stored. Such an initial value or subtraction value of the timer value may be adjusted based on the actual results. In this case, as shown in FIG. 7, a log management server 50 connected to each process apparatus via a network is provided. The log management server 50 includes a log acquisition unit 51, a log storage unit 52, and a log evaluation unit 53 that analyzes a log.

  The log acquisition unit 51 acquires log information about job execution from each process device via the network and accumulates it in the log storage unit 52. This log information includes a transaction identifier, a process device identifier, a process identifier, a timer value, an actual time, and timeout information.

The transaction identifier is an identifier for specifying each transaction request. The process device identifier is an identifier for identifying a process device that has executed processing in response to each transaction request. The process identifier is an identifier for specifying a process executed in the process device in response to each transaction request. The timer value is a timer value set in the transaction request or processing request. The actual time is the time required to respond to a transaction request or a processing request. The timeout information is information for determining whether timeout has occurred. With the transaction identifier and the device identifier, it is possible to specify the cooperative processing executed by each process device for one transaction request.

The log evaluation unit 53 analyzes the log information stored in the log storage unit 52. The log evaluation unit 53 includes an evaluation target memory that temporarily stores cooperation processing information to be alerted.
The log evaluation process executed in the log management server 50 will be described with reference to FIG. This log evaluation process is executed based on an instruction from an administrator terminal connected to the log management server 50.

  First, the log management server 50 executes a timeout situation search process (step S5-1). Specifically, the log evaluation unit 53 of the log management server 50 searches the log storage unit 52 for log information in which timeout information is recorded.

  Here, when the log in which the time-out information is recorded is extracted from the log storage unit 52, the log management server 50 executes the cooperation information acquisition process until time-out (step S5-2). Specifically, the log evaluation unit 53 of the log management server 50 extracts the log information in which the transaction identifier of the log information in which the timeout is recorded is recorded from the log storage unit 52. Using the process identifier stored in the log information, the cooperative processing executed before the timeout is specified. Then, the log evaluation unit 53 temporarily records the extracted cooperation processing information in the evaluation target memory. Note that if the log in which the timeout information is recorded is not extracted, the log management server 50 skips the cooperation information acquisition process (step S5-2) until the timeout.

  Next, the log management server 50 executes a transaction request search process in which there is little time before timeout (step S5-3). Specifically, the log evaluation unit 53 of the log management server 50 calculates the remaining time from when the transaction response message is received to the initial timer value in the log information recorded in the log storage unit 52. Then, when there is log information whose remaining time is shorter than a predetermined reference time, the log evaluation unit 53 determines that there is a transaction request with a small time margin.

  Here, when the transaction request with a small time margin until the timeout is extracted, the log management server 50 executes the acquisition process of the cooperation information regarding the transaction request (step S5-4). Specifically, the log evaluation unit 53 of the log management server 50 extracts, from the log storage unit 52, log information in which a transaction identifier of a transaction request having a small time margin is recorded. Using the process identifier stored in the log information, the cooperative process executed until the transaction response message is received is specified. Then, the log evaluation unit 53 temporarily records the extracted cooperation processing information in the evaluation target memory. In addition, when not extracting the transaction request with little time margin until timeout, the log management server 50 skips the acquisition process (step S5-4) of the cooperation information regarding a transaction request.

  Next, the log management server 50 executes determination processing as to whether or not alerting is necessary (step S5-5). Specifically, the log evaluation unit 53 of the log management server 50 determines that alerting is necessary when the cooperation processing information is stored in the evaluation target memory.

When it is determined that alerting is not necessary (in the case of “NO” in step S5-5), the log management server 50 ends the log evaluation process.
On the other hand, if it is determined that alerting is necessary (in the case of “YES” in step S5-5), the log management server 50 executes a process for calculating the time required for each process (step S5-6). Specifically, the log evaluation unit 53 of the log management server 50 calculates the actual time for each process identifier in the log information stored in the evaluation target memory. And the log evaluation part 53 calculates the statistical value (for example, average value) of performance time for every process.

  Next, the log management server 50 executes a specific process for a process having a long processing time (step S5-7). Specifically, the log evaluation unit 53 of the log management server 50 rearranges the calculated time statistics for each calculated process in descending order.

  Next, the log management server 50 executes output processing of cooperation status information (step S5-8). Specifically, the log evaluation unit 53 of the log management server 50 specifies the cooperation process in a transaction request including a process with a long processing time. And the log evaluation part 53 acquires the customer identifier and transaction classification identifier in this cooperation process from a transaction information storage part, and outputs them to an administrator terminal.

  As a result, the administrator can reinforce the device capability of the process whose processing time is long, or can review the initial value or subtraction value of the timer value for suppressing timeout.

  A to E: process device, A1 to E1 ... process execution unit, 10 to 40 ... timer monitoring unit, TM ... timer, 111 to 411 ... timer value setting means, 112 to 412 ... event monitoring means, 113 to 413 ... timeout report means.

Claims (6)

In accordance with the first message received from the preceding process, a process execution unit for executing a takeover process for transmitting a second message for requesting the subsequent process to be processed and receiving a return message from the subsequent process;
Connected to the timer to notify the timer value
A timeout monitoring system comprising a timer monitoring unit that acquires event information from the process execution unit and the timer,
The timer monitoring unit is
Obtaining a first timer value from the first message received by the process execution unit;
Calculating a second timer value obtained by subtracting an estimated time required for the takeover process from the first timer value;
A timer value setting means for setting the second timer value in a second message to be taken over by the subsequent process in the process execution unit;
Monitoring means for waiting for elapse of the second timer value in the timer and a return message from the subsequent process in the process execution unit;
If a timeout of the second timer value is detected before receiving the return message, a timeout report is output,
A timeout monitoring system comprising: a timeout report unit that allows the process execution unit to perform processing based on the return message when the return message is received before the second timer value times out.   The timeout monitoring system according to claim 1, wherein the timer monitoring unit holds a predetermined value in advance as an estimated time required for the takeover process. The timer monitoring unit
Based on the return message received in the process execution unit, determine the necessity of the subsequent linkage process,
When a new linkage process is performed, a remaining time obtained by subtracting an elapsed time until the return message is received from the first timer value is calculated as a second timer value;
3. The timeout monitoring system according to claim 1, wherein the process execution unit sets the second timer value in a second message to be taken over by a new linkage process. 4. The timer monitoring unit
If there is a subsequent linkage process, compare the remaining time for the subsequent linkage process with the reference value,
4. The timeout monitoring system according to claim 3, wherein a timeout report is output when the remaining time is equal to or less than a reference value.   The said timer monitoring part sets the initial value of a timer value based on the content of a transaction request, when receiving a transaction request for the first time, As described in any one of Claims 1-4 characterized by the above-mentioned. Timeout monitoring system. In accordance with the first message received from the preceding process, a process execution unit for executing a takeover process for transmitting a second message for requesting the subsequent process to be processed and receiving a return message from the subsequent process;
Connected to the timer to notify the timer value
A program for monitoring a timeout using a timeout monitoring system including a timer monitoring unit that acquires event information from the process execution unit and the timer,
The timer monitoring unit,
Obtaining a first timer value from the first message received by the process execution unit;
Calculating a second timer value obtained by subtracting an estimated time required for the takeover process from the first timer value;
Timer value setting means for setting the second timer value in a second telegram to be taken over by the subsequent process in the process execution unit;
Monitoring means for waiting for the elapse of the second timer value in the timer and a return message from the subsequent process in the process execution unit;
If a timeout of the second timer value is detected before receiving the return message, a timeout report is output,
When the return message is received before the second timer value times out, the process execution unit causes the process execution unit to function as a timeout report unit that permits processing based on the return message.

Images