JP2017228007A - Service continuation device, service continuation method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To continue a service even if a fault occurs in a module in a system which provides the service by combining divided multiple modules.SOLUTION: A service continuation device 10 comprises: a storage part 13 for storing a history record DB 130 on which a call source module, a call destination module, a call content, a response content and a score indicating the number of times of responses are recorded; a history recording part 112 which receives a response from the call destination module, uses history information in the call to determine whether or not to record the history information on the basis of a recording determination logic, and records the history information determined to record on the history recording DB; and a substitutive response part 115 for referring to the history recording DB when a call for which a fault detected module becomes a call destination module is received, and returning the response content on the history recording DB to the call source module if matched history information is recorded and the score is equal to or higher than a predetermined threshold.SELECTED DRAWING: Figure 2

Description

  The present invention relates to a service continuation apparatus, a service continuation method, and a program for providing a service by an application combining modularized software.

As the demand for speeding up software development increases, it is becoming common to divide software into small units and modularize them to create applications by combining those modularized software.
For example, Non-Patent Document 1 discloses a method of constructing a single application as a combination of small service groups. However, with this method, since the application is composed of a plurality of modules, there is a problem that a failure of one module affects the entire application.

  In order to solve this problem, a technique using a circuit breaker described in Non-Patent Document 2 is known. This technology prepares a module (circuit breaker) to mediate between modules, periodically monitors the aliveness of each module, and when a malfunction such as no response is detected, a call to the module in which the malfunction occurred occurs. In this case, the circuit breaker performs a NG response on behalf of the circuit breaker.

James Lewis, “Microservices,” [online], 25 March 2014, [Search June 8, 2016], Internet <URL: http://martinfowler.com/articles/microservices.html> Martin Fowler, “CircuitBreaker,” [online], 6 March 2014, [Search June 8, 2016], Internet <URL: http://martinfowler.com/bliki/CircuitBreaker.html>

  In the circuit breaker technology described in Non-Patent Document 2, when a malfunction of a module is discovered, an NG response is made on behalf of the surrogate, so that it is possible to prevent the influence of the malfunction from being propagated to the whole (fail-safe). . For example, by waiting for a response when a certain module calls a malfunctioning module, it is possible to avoid the event that the processing of other modules is delayed and the service falls into an unexpected error. However, with this circuit breaker technology, it is difficult to continue the service because it simply performs an NG response.

  In order to prevent the service from being stopped, it is possible to record the contents of the call to the module and the response contents from the module and return the recorded response contents if the call contents are the same. However, depending on the state of the module, different responses may be returned even for the same call contents. In the case of a call to a module having such a state (stateful module), there is a possibility that different responses are returned with the same call content, so the call content and the response content are simply recorded and recorded. There is a possibility of returning an incorrect response just by returning the response content.

  The present invention has been made in view of the above points, and the present invention provides a service in a system that provides a service by combining a plurality of divided modules even when a failure occurs in a certain module. An object is to provide a service continuation apparatus, a service continuation method, and a program that can be used.

  In order to solve the above-mentioned problem, the invention according to claim 1 is a system that provides a service by combining a plurality of divided modules, and is a call source module indicating the module that is a call source and a call destination. A service continuation device that mediates call and response processing with a callee module indicating the module, the caller module, the callee module, call contents attached to a call from the caller module; A history record DB (DataBase) in which the response contents attached to the response from the call destination module and the score indicating the number of times the response contents are responded to the call are recorded, and the response is executed on behalf A storage unit for storing a score threshold for determining whether or not to use the score, and from the caller module A call content notification unit that receives the call with the call content and notifies the call destination module; and receives a response to the call from the call destination module; the call source module, the call destination module, the call content; And a history recording unit that determines whether or not to record the history information based on a recording determination logic using history information indicating the response content, and records the history information determined to be recorded in the history recording DB; A life / death monitoring unit that performs life / death monitoring of the module by detecting a failure of the plurality of modules, and when the module that has detected the failure receives a call to be the callee module, Refer to the history record DB using the caller module, the callee module, and the call contents, and match It is a case where it is determined whether there is history information, and matching history information is recorded. When the score of the history information is equal to or higher than the score threshold, the history information is indicated by the history information in the history recording DB. And a proxy response unit that returns the response content to the call source module, and the record determination logic is configured to provide the call source module, the call destination module, and the call content for the call received from the call source module. When the response content matches the history information recorded in the history record DB, the call source module, the call destination module, the call content, and the response content are recorded in the history record DB. The score is incremented, and the call source module, the call destination module, and the call content match. Is a logic that, when the response contents do not match, the call source module, the call destination module, the call contents, and the response contents for the call are not recorded in the history record DB. A service continuation device characterized by

  According to a fourth aspect of the present invention, in a system that provides a service by combining a plurality of divided modules, a call source module that indicates the module that is a call source, and a call destination module that indicates the module that is a call destination A service continuation method of a service continuation device that mediates the processing of calls and responses between the service continuation device, wherein the service continuation device is a call attached to a call from the call source module, the call destination module, and the call source module. A history record DB in which contents, a response content attached to a response from the callee module, and a score indicating the number of times the call has been answered by the response content are recorded, and the response is executed on behalf of Is stored in the storage unit, and the caller module is determined using the score. Receiving the call to which the call content is attached from the caller and notifying the callee module; receiving a response to the call from the callee module; the caller module, the callee module, the call content; And determining whether or not to record the history information based on recording determination logic using history information indicating the response content, and recording the history information determined to be recorded in the history recording DB; Monitoring the life of the module by detecting a failure of the module, and when the module detecting the failure receives a call that is the callee module, the caller module for the call, Refer to the history record DB using the call destination module and the call content, and match It is a case where it is determined whether there is history information, and matching history information is recorded. When the score of the history information is equal to or higher than the score threshold, the history information is indicated by the history information in the history recording DB. Returning the response content to the caller module, and the record determination logic is configured to provide the caller module, the callee module, the call content for the call received from the caller module, When the response content matches the history information recorded in the history record DB, the call source module, the call destination module, the call content, and the response content are recorded in the history record DB. As an example, the score is incremented, and the call source module, the call destination module, and the call content are matched. Is a logic that, when the response contents do not match, the call source module, the call destination module, the call contents, and the response contents for the call are not recorded in the history record DB. A service continuation method characterized by

  According to a fifth aspect of the present invention, in a system that provides a service by combining a plurality of divided modules, a call source module that indicates the module that is a call source, and a call destination module that indicates the module that is a call destination A computer as a service continuation device that mediates processing of calls and responses between the call source module, the call destination module, call contents attached to a call from the call source module, and a response from the call destination module And a history record DB in which the response content attached to the call and a score indicating the number of times the response has been answered to the call are recorded, and whether or not the response is executed on behalf of the call using the score Means for storing a score threshold for determination, a call to which the call content from the call source module is attached Means for receiving and notifying the callee module, receiving a response to the call from the callee module, and using the caller module, the callee module, the call content, and history information indicating the response content, Based on the record determination logic, it is determined whether or not to record the history information, the history information determined to be recorded is recorded in the history record DB, and the module life or death is detected by detecting a failure of the plurality of modules. Means for monitoring, when the module detecting the failure receives a call to be the call destination module, the call history module using the call source module, the call destination module, and the call content for the call; To determine whether there is matching history information, and the matching history information is recorded. And when the score of the history information is equal to or higher than the score threshold, the response content indicated by the history information in the history record DB is functioned as a means for returning to the caller module, The record determination logic is such that the call source module, the call destination module, the call content, and the response content for the call received from the call source module match the history information recorded in the history record DB. The call source module, the call destination module, the call content, and the response content are recorded in the history record DB, the score is incremented, and the call source module, the call destination module, and If the call content matches but the response content does not match, The call source module, the call destination module, the call content, and the response content are logic programs that do not record in the history record DB.

  In this way, according to the service continuation device or the like, history information with different response contents is recorded in the history record DB even if the call source module, the call destination module, and the call contents match, according to the record determination logic. You can avoid recording. That is, the response from the stateful module is not recorded in the history record DB. On the other hand, when a failure occurs in a module that does not have a state (stateless module), it is possible to respond by proxy using the response content recorded in the history record DB of the service continuation device, and the service is continued. It becomes possible.

  Further, in the invention according to claim 2, in the storage unit, the response content indicated by the history information recorded in the history record DB by the proxy response unit in association with the call source module is stored in the call source. A billing DB in which the number of responses returned to the module is recorded is further stored, and when the proxy response unit receives a call in which the module in which the failure is detected becomes the callee module, the history record DB When the response content indicated by the recorded history information is returned to the calling source module, the charging processing unit further includes a charging processing unit that records the number of responses to the calling DB for each calling source module. The service continuation device according to claim 1 is used.

  As described above, according to the service continuation apparatus, the number of times of response can be recorded in the charging DB instead of the call destination module in which the failure has occurred. Therefore, the administrator of each module can be charged according to the number of responses recorded in the charging DB.

  According to a third aspect of the present invention, the system includes a plurality of spare modules having the same function corresponding to each of the plurality of modules, and the proxy response unit of the service continuation apparatus detects the failure. When the detected module receives a call that becomes the call destination module, the call history module DB is referred to by using the call source module, the call destination module, and the call content for the call, and the matching history information If the matching history information is not recorded, and if the matching history information is recorded, but the score of the history information is less than the score threshold, Notify the standby module corresponding to the called module, receive a response from the standby module, and attach the response to the response The response content was service continuity apparatus of claim 1, characterized in that return to the calling module.

  In this way, according to the service continuation apparatus, even if the call source module, the call destination module, and the call content match, the history information having different response contents is called to the standby module by the record determination logic. To obtain a response and return it to the calling module. Therefore, even when a failure occurs in a stateful module, it becomes possible to continue the service reliably.

  According to the present invention, in a system that provides a service by combining a plurality of divided modules, a service continuation device, a service continuation method, and a program are provided that continue a service even when a failure occurs in a certain module. be able to.

It is a figure which shows the structure of the service system containing the service continuation apparatus which concerns on this embodiment. It is a functional block diagram which shows the structural example of the service continuation apparatus which concerns on this embodiment. It is a figure which shows the data structure of the log | history information recorded on log | history recording DB which concerns on this embodiment. It is a flowchart which shows the flow of the process set by a recording determination logic. It is a sequence diagram which shows the process of the service system containing the service continuation apparatus which concerns on this embodiment. It is a functional block diagram which shows the structural example of the service continuation apparatus which concerns on the modification 1 of this embodiment. It is a figure which shows the data structure of the billing information recorded on billing DB which concerns on the modification 1 of this embodiment. It is a figure which shows the structure of the service system containing the service continuation apparatus which concerns on the modification 2 of this embodiment. It is a functional block diagram which shows the structural example of the service continuation apparatus which concerns on the modification 2 of this embodiment. It is a sequence diagram which shows the process of the service system containing the service continuation apparatus which concerns on the modification 2 of this embodiment.

  Next, the service continuation device 10 and the like in a mode for carrying out the present invention (hereinafter referred to as “the present embodiment”) will be described.

<Overview>
First, an outline of processing executed by the service continuation apparatus 10 according to the present embodiment (see FIGS. 1 and 2 to be described later) will be described.
When the service continuation apparatus 10 according to the present embodiment executes a process by a plurality of modules and provides a service, the service continuation apparatus 10 executes a process that mediates between the modules. Specifically, transmission / reception regarding the call / response is not performed directly between the modules but via the service continuation apparatus 10. Further, the service continuation apparatus 10 determines whether or not to record the information on the call contents and the response contents between the modules as a cache according to the record determination logic (see FIG. 4 to be described later), and records the information as history information ( “History recording DB (DataBase) 130” in FIG. 3 to be described later). This record determination logic is referred to as a module having a state “hereinafter referred to as a“ stateful module ”. ) And a module having no state (hereinafter referred to as a “stateless module”), and records history information regarding a call / response from the stateless module.
Then, when the service continuation device 10 detects the occurrence of a failure for a certain module, and when there is a call for that module, the service continuation apparatus 10 determines whether or not the same call content is recorded in the recorded history information. When the information is confirmed and recorded, the response content indicated by the history information is returned when a predetermined condition described later (“score is equal to or greater than a predetermined threshold”) is satisfied.

  By doing so, according to the service continuation device 10 according to the present embodiment, even if a stateful module and a stateless module are mixed, the stateful module is determined by making a determination based on the recording determination logic. It is possible to increase the possibility of recording the response content from the stateless module without recording the response content from. Therefore, even when a failure occurs in a module, the service can be continued without returning an erroneous response content.

<System configuration>
Next, the configuration of a service system including the service continuation device 10 according to the present embodiment will be described.
FIG. 1 is a diagram illustrating a configuration of a service system including a service continuation apparatus 10 according to the present embodiment.
In the service system according to the present embodiment, a plurality of data centers 100 (“1”... “N”) are connected via a network. Each data center 100 is provided with a server 20. The server 20 includes a module 2 in which software is divided into small units. In FIG. 1, servers “1” and “2” are installed in a data center “1”, the server “1” includes modules “1” and “2”, and the server “2” includes module “3”. The example provided with is shown. In the following, the symbol “2” is added as a generic term for the module 2, and when identifying the individual module 2, they are represented as module “1”, module “2”, module “3”, and the like.

  A service continuation device 10 is provided in the service system. This service continuation apparatus 10 is connected to each server 20 via a network. In FIG. 1, the service continuation apparatus 10 is illustrated as being installed in the data center “1”, but is not limited thereto, and may be installed in any other data center 100 or data It may be installed alone outside the center 100 after being connected to a network.

  Service provision in this service system is performed by combining the processing of the modules 2 in each server 20. At that time, transmission / reception regarding the call / response among the plurality of modules 2 is performed via the service continuation apparatus 10.

The service continuation apparatus 10 includes a module mediation unit 11 that mediates call / response between a plurality of modules 2, and records information about call contents and response contents between modules according to a record determination logic (see FIG. 3). Keep a record. Then, when there is a call to the module 2 in which the failure has occurred, the service continuation apparatus 10 refers to the history record DB 130 and determines whether or not the same call content is recorded in the recorded history information. If the predetermined condition (“score is equal to or higher than a predetermined threshold”) is satisfied, the response content indicated by the history information is returned.
Hereinafter, the service continuation device 10 will be described in detail.

<Configuration of service continuation device>
Next, the configuration of the service continuation device 10 according to the present embodiment will be described.
FIG. 2 is a functional block diagram illustrating a configuration example of the service continuation apparatus 10 according to the present embodiment.
As illustrated in FIG. 2, the service continuation device 10 includes a control unit (not shown) including the function of the module mediation unit 11, an input / output unit 12, and a storage unit 13.

  The input / output unit 12 performs input / output of information with the modules 2 and the like provided in each server 20 that are communicatively connected via a network. The input / output unit 12 performs input / output between a communication interface (not shown) that transmits and receives information via a communication line and an input means such as a keyboard and an output means such as a monitor (both not shown). I / O interface to perform.

The storage unit 13 includes a hard disk, a flash memory, a RAM (Random Access Memory), and the like.
The storage unit 13 stores a history record DB 130 (FIG. 3), a score threshold 140, and the like, which will be described later, a program for executing each function of the control unit (module mediation unit 11), Information necessary for processing is temporarily stored.

The control unit (module intermediary unit 11) controls the service continuation device 10 as a whole, and includes a call content notification unit 111, a history recording unit 112, a response content return unit 113, a life / death monitoring unit 114, and a proxy response unit 115. It is comprised including.
The control unit is realized, for example, when a program (CPU) (not shown) expands and executes a program stored in the storage unit 13 on a RAM.

The call content notification unit 111 receives a request (call with call content) from a certain module 2 (calling module) to another module 2 (calling module). Then, the call content notification unit 111 notifies the request (call) to the module 2 that is the call destination.
Further, when the call content notification unit 111 receives a request (call) from the call source module to the call destination module, the call content notification unit 111 receives information from the alive monitoring unit 114 that a failure has occurred in the call destination module. If so, the request (call) is output to the proxy response unit 115.

  The history recording unit 112 receives a response from the callee module to which the call content notification unit 111 has notified the request. In order to link which request (call) is the response to this response from the callee module, information identifying the request (call) such as the sequence number and the ID / call contents of the caller module, It is included in the response content. Then, the history recording unit 112 specifies a request (call) corresponding to a response from the callee module, and in accordance with the record determination logic, information including the response content attached to the response is used as history information in the history recording DB 130. It is determined whether or not to record. When the history recording unit 112 determines to record, the history recording unit 112 records the history information in the history recording DB 130 illustrated in FIG. 3 as history information.

FIG. 3 is a diagram illustrating a data configuration of history information recorded in the history recording DB 130 according to the present embodiment.
As shown in FIG. 3, the history record DB 130 stores each item of the call source module 131, the call destination module 132, the call content 133, the response content 134, and the score 135 as history information.

The caller module 131 stores the identifier of the module 2 that is the caller of the request received by the service continuation apparatus 10. For example, as shown in FIG. 3, the identifier “3” of the module 2 that is the call source is stored in the call source module 131.
The call destination module 132 stores the identifier of the module 2 that is the call destination of the received request. For example, as shown in FIG. 3, the identifier “1” of the module 2 that is the call destination is stored in the call destination module 132.

The call content 133 stores the call content attached to the request from the call source module. For example, as shown in FIG. 3, “GET / index.html” is stored as the call content 133.
The response content 134 stores the response content attached to the response from the callee module.
For example, as shown in FIG. 3, “HTTP / 1.1 200OK hoge” is stored as the response content 134.

The score 135 includes a history of each information of the call source module 131, the call content 133 attached to the request from the call source module, the call destination module 132, and the response content 134 attached to the response from the call destination module. The number of times that the recording unit 112 determines that the history information is the same based on the recording determination logic is stored.
In the example shown in FIG. 3, the call source module 131 is “3”, the call destination module 132 is “1”, the call content 133 is “GET / index.html”, and the response content 134 is “HTTP / 1.1 200OK hoge”. This indicates that the process was “323 (times)”.

Next, the record determination logic used by the history recording unit 112 to determine whether or not to record in the history recording DB 130 will be described.
FIG. 4 is a flowchart showing the flow of processing set by the record determination logic.
The history recording unit 112 includes a call source module and call content information acquired by the call content notification unit 111 upon notification of a request to the call destination module, and a call destination module attached to a response received from the call destination module. When the response is received based on the response content information, the following processing (steps S1 to S5) based on the record determination logic is executed.

First, the history recording unit 112 records, in the history recording DB 130 (FIG. 3), all of these pieces of information regarding the call source module, the call destination module, and the call content related to the response received from the call destination module ( It is determined whether there is (history information) (step S1).
Then, when there is no record in which all of these pieces of information match (step S1 → No), the history recording unit 112 creates a new record (history information) as a callee module, a caller module, call contents, and The response content is stored in the history recording DB 130, and the score 135 is set to "1" (step S2).
On the other hand, if there is a record that matches all in step S1 (step S1 → Yes), the history recording unit 112 proceeds to step S3.

  In step S <b> 3, the history recording unit 112 determines whether or not there is a record in the history record DB 130 that matches all of the call source module, the call destination module, the call content, and the response content. If there is no record that matches all (step S3 → No), the process proceeds to step S4.

  In step S4, the history recording unit 112 determines not to record the information to be determined as a record in the history recording DB 130 for the call source module, the call destination module, the call content, and the response content. In step S1, the history recording unit 112 executes a process of deleting the call source module, the call destination module, and the record whose call contents match from the history record DB 130.

  On the other hand, if there is a record that matches all in step S3 (step S3 → Yes), the history recording unit 112 increments the score 135 of the record (step S5). Then, the process of the record determination logic is finished.

  Returning to FIG. 2, when the record determination logic process by the history recording unit 112 is completed, the response content return unit 113 returns the response received from the callee module to the caller module.

The alive monitoring unit 114 monitors the aliveness of each module 2 by transmitting a alive confirmation request to each module 2 at a predetermined time interval and receiving the response information.
The life and death monitoring unit 114 outputs the failure occurrence information to the call content notification unit 111 when the occurrence of a failure for a certain module 2 is detected by the life and death monitoring of each module 2.

The proxy response unit 115 extracts the call source module and the call content from the request to the call destination module where the failure has occurred based on the output from the call content notification unit 111, and stores the call source module and the call content in the history record DB 130 (FIG. 3). Among them, it is confirmed whether or not a record having the same call source module, call destination module, and call content is recorded. Then, when the proxy response unit 115 is recorded, the proxy response unit 115 determines whether or not the score 135 of the record is equal to or higher than the score threshold value 140 preset in the storage unit 13. If the value of the score 135 is greater than or equal to the score threshold value 140, the proxy response unit 115 returns the response content 134 indicated by the record to the caller module.
On the other hand, the proxy response unit 115 determines that the value of the score 135 is the score 135 even if the call source module, the call destination module, and the record that matches the call content are not recorded in the history record DB 130. When it is less than the threshold value 140, an NG response is returned to the caller module.

≪Process flow≫
Next, a flow of processing executed by a service system including the service continuation apparatus 10 according to the present embodiment will be described.
FIG. 5 is a sequence diagram showing processing of the service system including the service continuation apparatus 10 according to the present embodiment. It is assumed that a score threshold value 140 (here, “300”) is registered in the storage unit 13 of the service continuation apparatus 10 in advance.

First, processing (steps S10 to S14) before a failure occurs in each module 2 (normal time) will be described.
As shown in FIG. 5, for example, the service continuation apparatus 10 receives a request (GET / index.html) from the module “3” (calling module) (step S10).
Subsequently, the call content notification unit 111 of the service continuation apparatus 10 notifies the request (GET / index.html) to the module “1” that is the call destination module (step S11).

Next, the service continuation apparatus 10 receives a response (HTTP / 1.1 200 OK hoge) from the module “1” that is the call destination module (step S12).
When the history recording unit 112 of the service continuation device 10 receives the response (HTTP / 1.1 200 OK hoge) from the callee module (module “1”), it is attached to the response according to the record determination logic shown in FIG. It is determined whether or not information including the response content is recorded in the history record DB 130 as history information (step S13).

  Here, in the processing of the record determination logic shown in FIG. 4, in step S1, the call source module, the call destination module, and the call content are module “3”, module “1”, (GET / index.html). Is recorded in the history record DB 130 (step S1 → Yes in FIG. 4), and in step S3 in FIG. 4, the response content also matches (HTTP / 1.1 200OK hoge). In the history record DB 130 (step S3 → Yes in FIG. 4). In this case, the history recording unit 112 increments the score 135 of the matching record (step S5 in FIG. 4). The state as a result of this increment will be described below assuming that it is the state shown in FIG.

  Next, the response content return unit 113 of the service continuation apparatus 10 returns the response (HTTP / 1.1 200 OK hoge) received from the call destination module (module “1”) to the call source module (step S14).

In the above normal processing, the service continuation apparatus 10 updates the history record DB 130 as needed in accordance with the call / response between the modules 2. In addition, the service continuation device 10 performs alive monitoring of each module 2.
The alive monitoring unit 114 of the service continuation apparatus 10 transmits a alive confirmation request to each module 2 at a predetermined time interval (step S20). Here, it is assumed that a failure has occurred in module “1”.
The life and death monitoring unit 114 receives response information for the life and death confirmation request from the modules “2” and “3” (step S21), but cannot receive response information from the module “1”. Thereby, the life and death monitoring unit 114 detects that a failure has occurred in the module “1” (step S22). Then, the life and death monitoring unit 114 outputs to the call content notification unit 111 that a failure has occurred in the module “1”.

In this state, the service continuation apparatus 10 receives a request (GET / index.html) from the module “3” (caller module) to the module “1” (step S30).
In this case, the call content notification unit 111 of the service continuation apparatus 10 outputs a request to the proxy response unit 115 because a failure has occurred in the module “1”.

The proxy response unit 115 of the service continuation apparatus 10 uses the transmission source module (module “3”), transmission destination module (module “1”), and call content (GET / index.html) attached to the request. Then, the history record DB 130 is searched and it is confirmed whether or not a record having the same information is recorded (step S31).
In the case where the record is recorded, the proxy response unit 115 determines whether or not the score 135 of the record is equal to or greater than the score threshold 140 (“300”).
Here, since the value of the score 135 of the record shown in the first line of the history record DB 130 shown in FIG. 3 is “323”, the proxy response unit 115 determines that the score is greater than or equal to the score threshold 140 (“300”). .
Then, the proxy response unit 115 returns the response content 134 (here, “HTTP / 1.1 200OK hoge”) recorded in the record to the transmission source module (module “3”) (step S32: proxy response). .

Next, the service continuation apparatus 10 receives a request (GETID user1) from the module “2” (caller module) to the module “1” (step S40).
Also in this case, the call content notification unit 111 of the service continuation apparatus 10 outputs the request to the proxy response unit 115 because a failure has occurred in the module “1”.

The proxy response unit 115 of the service continuation apparatus 10 records the history using the transmission source module (module “2”), transmission destination module (module “1”), and call contents (GETID user1) attached to the request. The DB 130 is searched, and it is confirmed whether or not a record matching these pieces of information is recorded (step S41).
In the case where the record is recorded, the proxy response unit 115 determines whether or not the score 135 of the record is equal to or greater than the score threshold 140 (“300”).
Here, since the value of the score 135 of the record shown in the second line of the history record DB 130 shown in FIG. 3 is “5”, the proxy response unit 115 determines that the score is not equal to or higher than the score threshold 140 (“300”). .
Then, the proxy response unit 115 returns an NG response to the transmission source module (module “2”) (step S42).

  By doing in this way, according to the service continuation apparatus 10 etc. which concern on this embodiment, the response content from the stateless module 2 can be reliably recorded on log | history recording DB130 by a recording determination logic. Therefore, when a failure occurs in the stateless module 2, the service continuation device 10 can respond on behalf of the service, so that the service can be continued.

<Modification 1 of this embodiment>
Next, Modification 1 of the service continuation apparatus 10 according to the present embodiment will be described.
The service continuation device 10A (10) according to the first modification of the present embodiment includes a billing DB 150 (see FIGS. 6 and 7 to be described later). Then, when the service continuation device 10A (10) uses the information recorded in the history record DB 130 to respond to the transmission source module as a proxy, the service continuation device 10A (10) records in the billing DB.

FIG. 6 is a functional block diagram illustrating a configuration example of the service continuation device 10A (10) according to the first modification of the present embodiment.
Compared with the service continuation apparatus 10 according to the present embodiment shown in FIG. 2, the storage unit 13 includes a charging DB 150, and the control unit (module mediation unit 11) includes a charging processing unit 116. In addition, about the structure provided with the same function as the structure shown in FIG. 2, the same name and code | symbol are attached | subjected and description is abbreviate | omitted.

  The billing processing unit 116 monitors the processing of the proxy response unit 115, and the proxy response unit 115 searches the history record DB 130 for a call destination module in which a failure has occurred, and history information having a score 135 value equal to or higher than the score threshold 140. Is recorded in the charging DB 150 when the response content is returned by proxy.

FIG. 7 is a diagram showing a data configuration of billing information recorded in the billing DB 150 according to the first modification of the present embodiment.
As shown in FIG. 7, each item of the caller module 151, the administrator ID 152, and the number of responses 153 is stored in the charging DB 150 as charging information.

The caller module 151 stores the identifier of the caller module 2 of the request received by the service continuation apparatus 10. This caller module 151 is the same information as the caller module 131 in the history record DB 130 of FIG.
The administrator ID 152 stores an identifier of an administrator who manages the caller module 151. For example, “# 0003” is stored as the manager ID 152 of the manager who manages the module 2 whose call source module 151 is “3”.

  The number of responses 153 stores the number of times that the proxy response unit 115 responded using the response content 134 recorded in the history record DB 130 instead of the call destination module in which the failure occurred. For example, in the example illustrated in FIG. 7, a response of “45 (times)” is stored in response to the call source module 151 instead of the call destination module in which the failure has occurred.

  As described above, according to the service continuation device 10A (10) according to the first modification of the present embodiment, the proxy response unit 115 records the number of times of responding instead of the failed call destination module in the accounting DB 150. Can do. Therefore, the administrator of each module 2 can be charged according to the response count 153 recorded in the charging DB 150 of the service continuation apparatus 10A (10).

<Modification 2 of this embodiment>
Next, a second modification of the service continuation device 10 according to the present embodiment will be described.
FIG. 8 is a diagram illustrating a configuration of a service system including the service continuation device 10B (10) according to the second modification of the present embodiment.
The service system shown in FIG. 8 is characterized in that a spare module 2B (2) is provided in the data center “2” in addition to the service system shown in FIG. The spare module 2B (2) is a spare module corresponding to each of the modules 2 included in the data center “1” and the like. That is, the module 2 and the spare module 2B (2) have a function of executing the same processing. FIG. 8 illustrates an example in which spare modules “1”, “2”, and “3” corresponding to the modules “1”, “2”, and “3” are provided in the server “3” of the data center “2”. However, each of the spare modules 2B (2) is not limited to being provided in one data center 100, and may be provided in a plurality of data centers 100. Further, as shown in FIG. 8, it is not limited to storing all spare modules 2B (2) in one server 20 (in this case, server “3”), but distributed to a plurality of servers 20. It may be stored.

  The service continuation device 10B (10) according to the second modification of the present embodiment includes a module mediation unit 11B (11) as illustrated in FIG. This module mediation unit 11B (11) has a proxy response unit 115B (115) different from the proxy response unit 115 of the service continuation apparatus 10 shown in FIG. In addition, about the structure provided with the same function as the structure shown in FIG. 2, the same name and code | symbol are attached | subjected also in FIG. 9, and description is abbreviate | omitted.

  The proxy response unit 115B (115) extracts the call source module and the call content based on the request for the call destination module in which the failure has occurred, similarly to the proxy response unit 115 of the present embodiment shown in FIG. In the history record DB 130 (see FIG. 3), it is confirmed whether or not a record having the matching call source module, call destination module, and call content is stored. Then, if the proxy response unit 115B (115) is recorded, the proxy response unit 115B (115) determines whether or not the score 135 of the record is equal to or higher than the score threshold 140 preset in the storage unit 13. If the value of the score 135 is greater than or equal to the score threshold value 140, the proxy response unit 115 returns the response content 134 indicated by the record to the caller module.

  On the other hand, the proxy response unit 115B (115) has a score of 135 even if the call record module, the call destination module, and a record with matching call contents are not recorded or recorded in the history record DB 130. When the value is less than the score threshold 140, a request is transmitted to the backup module 2B (2) corresponding to the call destination module in which the failure has occurred. Then, the proxy response unit 115B (115) acquires a response from the spare module 2B (2) corresponding to the call destination module, and returns the response to the call source module.

≪Process flow≫
Next, a flow of processing executed by a service system including the service continuation device 10B (10) according to the second modification of the present embodiment will be described.
FIG. 10 is a sequence diagram showing processing of the service system including the service continuation device 10B (10) according to the second modification of the present embodiment. It is assumed that the score threshold 140 (here, “300”) is registered in the storage unit 13 of the service continuation device 10B (10). Further, the service continuation device 10B (10) updates the history record DB 130 as needed in accordance with the call / response between the modules 2 by the normal processing shown in steps S10 to S14 in FIG. Then, it is assumed that the service continuation device 10B (10) detects the failure of the module “1” by executing the alive monitoring of each module 2 (steps S20 to S22). Here, it is assumed that the history information recorded in the history record DB 130 of the service continuation apparatus 10B (10) after the occurrence of the failure of the module “1” in step S22 is in the state shown in FIG.

After step S22, service continuation apparatus 10B (10) receives a request (GETID user1) from module “2” (calling source module) to module “1” (destination module) (step S50).
In this case, the call content notification unit 111 of the service continuation apparatus 10B (10) outputs a request to the proxy response unit 115B (115) because a failure has occurred in the module “1”.

  The proxy response unit 115B (115) of the service continuation apparatus 10B (10) adds the transmission source module (module “2”), the transmission destination module (module “1”), and the call content (GETID user1) to the request. Is used to search the history record DB 130 (FIG. 3), and it is confirmed whether or not a record having the same information is recorded (step S51).

If the response is recorded, the proxy response unit 115B (115) determines whether or not the score 135 of the record is equal to or greater than the score threshold 140 (“300”).
Here, since the value of the score 135 of the record shown in the second line of the history record DB 130 shown in FIG. 3 is “5”, the proxy response unit 115B (115) is not equal to or higher than the score threshold 140 (“300”). Is determined.

  Subsequently, the proxy response unit 115B (115) transmits a request to the spare module 2B (2) corresponding to the call destination module in which the failure has occurred (step S52). Here, the proxy response unit 115B (115) transmits a request (GETID user1) to the spare module “1” corresponding to the module “1” that is the call destination in which the failure has occurred.

Then, the proxy response unit 115B (115) of the service continuation device 10B (10) receives the response (“1018714”) from the backup module “1” (step S53).
Next, the proxy response unit 115B (115) returns the received response (“1018714”) to the caller module (module “2”) (step S54).

  By doing in this way, the service continuation device 10B (10) according to the second modification of the present embodiment is stateless by the recording determination logic even when the stateful module 2 and the stateless module 2 are mixed. The response contents from the module 2 can be reliably recorded in the history record DB 130. Therefore, when a failure occurs in the stateless module 2, the service continuation device 10B (10) can respond on behalf. Further, the service continuation device 10B (10) can obtain a response from the spare module 2B (2) and return it to the caller module when a failure occurs in the stateful module 2. Therefore, according to the service continuation device 10B (10) according to the second modification of the present embodiment, it is possible to continue the service more reliably when a failure occurs in the module 2.

2 module 2B spare module 10, 10A, 10B service continuation device 11, 11B module mediation unit 12 input / output unit 13 storage unit 100 data center 111 call content notification unit 112 history recording unit 113 response content return unit 114 alive monitoring unit 115, 115B Proxy response unit 116 Charge processing unit 130 History record DB
140 Score threshold 150 Billing DB

Claims (5)

In a system for providing a service by combining a plurality of divided modules, processing of a call and a response between a call source module indicating the module serving as a call source and a call destination module indicating the module serving as a call destination A service continuation device that mediates,
The caller module, the callee module, the call content attached to the call from the caller module, the response content attached to the response from the callee module, and responding to the call with the response content A storage unit for storing a history recording DB (DataBase) in which a score indicating the number of times is recorded, and a score threshold for determining whether to execute the response on behalf of the score, using the score;
A call content notification unit that receives a call with the call content from the call source module and notifies the call destination module;
A response to the call is received from the call destination module, and the history information is recorded based on a record determination logic using the call source module, the call destination module, the call content, and history information indicating the response content. A history recording unit that determines whether to record and records history information determined to be recorded in the history recording DB;
A life / death monitoring unit that performs life / death monitoring of the module by detecting a failure of the plurality of modules,
When the module that has detected the failure receives a call that becomes the call destination module, the call record module, the call destination module, and the call content for the call are referred to the history record DB to match. It is a case where there is history information to be matched, and when matching history information is recorded, and when the score of the history information is greater than or equal to the score threshold, the history information in the history recording DB A proxy response unit for returning the response content shown to the caller module;
The record determination logic is
When the call source module, the call destination module, the call content, and the response content for the call received from the call source module match the history information recorded in the history record DB, the call source Assuming that the module, the call destination module, the call content, and the response content are recorded in the history record DB, the score is incremented, and the call source module, the call destination module, and the call content match. However, if the response content does not match, the call source module, the call destination module, the call content, and the response content for the call are not recorded in the history record DB. A service continuation device characterized by being. The storage unit records the number of responses that the proxy response unit returns to the call source module the response content indicated by the history information recorded in the history record DB in association with the call source module DB is further memorized,
When the module that detected the failure receives a call that becomes the call destination module, the proxy response unit returns the response content indicated by the history information recorded in the history record DB to the call source module. 2. The service continuation apparatus according to claim 1, further comprising: a billing processing unit that records the number of times of response to the billing DB for each caller module. The system includes a plurality of spare modules having the same function corresponding to each of the plurality of modules.
The proxy response unit of the service continuation device is:
When the module that has detected the failure receives a call that becomes the call destination module, the call record module, the call destination module, and the call content for the call are referred to the history record DB to match. If there is no history information that matches, and the history information that matches is recorded, but the score of the history information is less than the score threshold In addition, the call is notified to the spare module corresponding to the callee module, a response is received from the spare module, and the response content attached to the response is returned to the call source module. The service continuation device described in 1. In a system for providing a service by combining a plurality of divided modules, processing of a call and a response between a call source module indicating the module serving as a call source and a call destination module indicating the module serving as a call destination A service continuation method for a mediation service continuation device,
The service continuation device comprises:
The caller module, the callee module, the call content attached to the call from the caller module, the response content attached to the response from the callee module, and responding to the call with the response content A history recording DB in which a score indicating the number of times is recorded, and a score threshold for determining whether to execute the response on behalf using the score are stored in the storage unit,
Receiving a call with the call content from the caller module and notifying the callee module;
A response to the call is received from the call destination module, and the history information is recorded based on a record determination logic using the call source module, the call destination module, the call content, and history information indicating the response content. Determining whether or not to record, and recording history information determined to be recorded in the history recording DB;
Performing alive monitoring of the module by detecting a failure of the plurality of modules;
When the module that has detected the failure receives a call that becomes the call destination module, the call record module, the call destination module, and the call content for the call are referred to the history record DB to match. It is a case where there is history information to be matched, and when matching history information is recorded, and when the score of the history information is greater than or equal to the score threshold, the history information in the history recording DB Returning the response content shown to the caller module;
The record determination logic is
When the call source module, the call destination module, the call content, and the response content for the call received from the call source module match the history information recorded in the history record DB, the call source Assuming that the module, the call destination module, the call content, and the response content are recorded in the history record DB, the score is incremented, and the call source module, the call destination module, and the call content match. However, if the response content does not match, the call source module, the call destination module, the call content, and the response content for the call are not recorded in the history record DB. A service continuation method characterized by being. In a system for providing a service by combining a plurality of divided modules, processing of a call and a response between a call source module indicating the module serving as a call source and a call destination module indicating the module serving as a call destination A computer as a service continuation device to mediate,
The caller module, the callee module, the call content attached to the call from the caller module, the response content attached to the response from the callee module, and responding to the call with the response content Means for storing a history record DB in which a score indicating the number of times is recorded, and a score threshold for determining whether or not to execute the response on behalf using the score;
Means for receiving a call with the call content from the call source module and notifying the call destination module;
A response to the call is received from the call destination module, and the history information is recorded based on a record determination logic using the call source module, the call destination module, the call content, and history information indicating the response content. Means for recording the history information determined to be recorded in the history recording DB;
Means for alive monitoring of the module by detecting a failure of the plurality of modules;
When the module that has detected the failure receives a call that becomes the call destination module, the call record module, the call destination module, and the call content for the call are referred to the history record DB to match. It is a case where there is history information to be matched, and when matching history information is recorded, and when the score of the history information is greater than or equal to the score threshold, the history information in the history recording DB Functioning as a means for returning the response content shown to the caller module;
The record determination logic is
When the call source module, the call destination module, the call content, and the response content for the call received from the call source module match the history information recorded in the history record DB, the call source Assuming that the module, the call destination module, the call content, and the response content are recorded in the history record DB, the score is incremented, and the call source module, the call destination module, and the call content match. However, if the response content does not match, the call source module, the call destination module, the call content, and the response content for the call are not recorded in the history record DB. A program to make it function as it is.

Images