JP2006146434A - Distributed application test method and distributed application test equipment - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem in a distributed application test, that a method of manually correcting a distributed application for test and testing the distributed application in a working system or a method of constructing an environment for test different from an environment for the actual operation and testing the distributed application is low in the reliability to the test results. <P>SOLUTION: When a distributed application(AP1) 1410 for a test is arranged, a distributed AP distribution part 1220 is notified of an execution mode 1420 and an automatic conversion rule definition information 1430, and a distributed AP execution information automatic converting part 1230 automatically changes distributed AP information, and an object name and object information are registered in a name table 3213 for a test in a naming service part 3210. Thus, it is possible to make an existing application and an application for a test coexist, and to easily realize the migration of the distributed application after the test. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a test of a distributed application (hereinafter abbreviated as “distributed AP”) created using a distributed object technology or a component technology. In particular, the present invention relates to a technique for performing a distributed AP test on a computer in actual operation.

  A distributed AP (distributed application) is an execution of an AP that realizes a target process by a plurality of applications (hereinafter abbreviated as “AP”) distributed and deployed on a plurality of computers operating together. Is the method.

  When a distributed AP deployed on a certain computer requests processing to a distributed AP deployed on another computer, a method of searching for the target distributed AP is the method of OMG (Object Management Group). A method of acquiring object reference information using a naming service (Naming Service) indicated in “Naming Service Specification” is common.

  As a method for acquiring object information from a naming service, a speed-up technique as described in Patent Document 1 has also been proposed.

In addition, when an existing distributed AP is changed, as a method for verifying the operation of the distributed AP, a test environment similar to the production environment is built and a test is performed, or information for executing the distributed AP In general, the test is performed by deploying in a production environment as a distributed AP different from the existing distributed AP.
Japanese Patent Laying-Open No. 2004-30486 (paragraphs 0007 to 0011, FIG. 1)

  However, the method of building and testing a test environment similar to the production environment requires a construction cost because it needs to be constructed separately from the production environment, and is completely the same as the production environment. There is a problem that it is difficult to prepare an environment (connection destination information, various port information, etc.).

  Moreover, in the method of changing the information for executing the distributed AP and deploying the test in the production environment as a distributed AP different from the existing distributed AP, the information for executing the distributed AP to be deployed is used for the test. In order to move to the production environment after the test is completed, a manual change operation is again required for the distributed AP that has confirmed the corner operation. For this reason, a human error may occur, and the reliability of the test is reduced. In order to avoid this, an object of the present invention is to realize a test using a production environment without changing information for executing a distributed AP.

  In order to achieve the above-mentioned problem, when deploying a distributed AP to be tested, the deployment tool is notified of the fact that it is for testing, so that runtime information for implementation of the distributed AP (implementation code or execution environment) Definition information, deployment method and deployment identifier including information on AP operation not directly included in the implementation code, or conversion rule and execution mode information in the main storage device used at the time of execution) are automatically converted, and object reference information is By managing and referring to each execution mode individually, the existing distributed AP and the distributed AP to be tested are allowed to coexist.

  By adding an execution mode (production mode / test mode: however, there are a variety of test modes, not a single type) to each distributed AP, it is the same as the distributed AP being executed on the distributed AP server in production It is possible to perform an operation test of a distributed AP having an object name of Further, it is possible to perform the test without manually changing the runtime information of the distributed AP to be tested for testing, and it is possible to easily shift to the production environment after the test is completed.

  Hereinafter, the best mode for carrying out the present invention will be described in detail with reference to the drawings.

[Configuration of distributed application test equipment]
FIG. 1 is a diagram showing an embodiment of a distributed application test apparatus (distributed object system) according to the present invention. The present embodiment is a system realized by a computer 1000, a computer 2000, and a computer 3000 connected by a network 100. Among these, the computer 3000 is configured to have a naming service unit 3210 not included in the computer 1000 and the computer 2000 in the main storage device 3200, and has a role different from that of the computer 1000 and the computer 2000.

  The computer 1000 includes a CPU (Central Processing Unit) 1100 and a main storage device 1200. On the main storage device 1200, a container 1210 that provides an execution environment for distributed APs, a distributed AP deployment unit 1220 that deploys distributed APs, A distributed AP runtime information automatic conversion unit 1230, a naming interface unit 1240, a remote call unit 1250, and a distributed object base unit 1260 that automatically convert distributed AP runtime information are arranged. Furthermore, distributed APs 1211, 1212 and 1213 are arranged in the container 1210, and a name table access control unit 1241 exists in the naming interface unit 1240.

  The “deployment” used in the description of the present embodiment means (1) + (preparing a distributed AP on a container) and (2) starting a distributed AP prepared on a container. The meaning of 2) or (1) is indicated. The “run-time information” is a deployment identifier of the distributed AP, definition information of the execution environment, a deployment identifier including information on an AP operation that is not directly included in the deployment method or the implementation code, or a main information used at the time of execution. Refers to conversion rule and execution mode information in the storage device.

  The computer 2000 includes a CPU 2100 and a main storage device 2200. On the main storage device 2200, a container 2210 that provides a distributed AP execution environment, a distributed AP deployment unit 2220 that deploys a distributed AP, and runtime information of the distributed AP. A distributed AP runtime information automatic conversion unit 2230, a naming interface unit 2240, a remote call unit 2250, and a distributed object base unit 2260 are arranged. Furthermore, distributed APs 2211, 2122, and 2213 are arranged in the container 2210, and a name table access control unit 2241 exists in the naming interface unit 2240.

  The computer 3000 includes a CPU 3100 and a main storage device 3200, and a naming service unit 3210 and a distributed object base unit 3260 are arranged on the main storage device 3200. The naming service unit 3210 has a name table 3211, and this table has a production name table 3212 and a test name table 3213.

  In this embodiment, an example of a form in which the computer 1000 and the computer 2000 and the computer 3000 connected to the network are separated is shown. However, a distributed application test apparatus is configured by allocating a plurality of elements to one computer. Also good. Hereinafter, each part will be briefly described.

[container]
The containers 1210 and 2210 provide a distributed AP execution environment. Then, after the distributed AP is deployed in this container, execution is started. In the example of FIG. 1, the distributed APs 1211, 1212, 1213, 2211, 2213, and 2213 execute the requested processes after being deployed in the container.

[Distributed AP deployment unit]
The distributed AP deployment units 1220 and 2220 deploy the specified distributed AP. And the function etc. which switch the process for deployment according to execution mode are also included.

[Distributed AP runtime information automatic conversion unit]
The distributed AP runtime information automatic conversion units 1230 and 2230 automatically convert the distributed AP runtime information when the execution mode is the test mode. When the execution mode is the production mode, the distributed AP is started with the runtime information as it is without calling the distributed AP runtime information automatic conversion units 1230 and 2230.

[Naming interface part]
The naming interface units 1240 and 2240 are interfaces for accessing the naming service unit 3210. Each naming interface unit includes a name table access control unit.

[Name table access control unit]
The name table access control units 1241 and 2241 have a function of accessing the name table 3211. However, the name table access control units 1241 and 2241 are not directly accessed, but the communication functions of the distributed object base units 1260 and 2260 described later are used.

[Remote call section]
The remote call units 1250 and 2250 are communication layers that send requests to remote objects. As an example of a known technique, there is an RMI (Remote Method Invocation) technique shown at http://java.sun.com/products/jdk/rmi/, and this function is used in the example of the present embodiment. .

[Distributed Object Platform]
The distributed object infrastructure units 1260, 2260, and 3260 are communication layers that send requests to distributed objects. Similar to the remote calling units 1250 and 2250, the RMI technique is applied as it is. The distributed object base units 1260, 2260, and 3260 provide an environment for executing distributed objects on each computer. When calling a function of a computer or an object of another computer, a subcontracting process is performed to communicate with another computer.

[Naming Service Department]
The naming service unit 3210 stores the object name and object reference information notified to the name table 3211 or searches the object reference information using the notified object name as a key. The naming service unit 3210 includes a name table 3211 described later. It is sufficient that one naming service unit 3210 is installed in the system, but the number of naming service units is not necessarily limited to one.

[Name table]
The name table 3211 stores object names and object reference information of objects on the system by separating the storage destination table depending on whether it is a distributed AP for production or a distributed AP for testing.

[Flow of processing until test execution of distributed AP]
FIG. 2 is a diagram showing a flow of processing until the distributed AP is test-executed in the present embodiment. When testing a distributed AP, it is not possible to simply load the distributed AP and start immediately. First, a distributed AP to be tested needs to be deployed. This deployment stage is further divided into two stages. The first stage is a deployment process performed by the distributed AP runtime information automatic conversion unit (S11). In the first stage, the distributed AP runtime information is converted, and the distributed AP deployment unit 1220 performs container conversion. To be prepared. The second stage is a deployment process performed by the name table access control unit (S12). By this process, the distributed AP whose runtime information is converted in the first stage can be started. Then, in order to call another distributed AP from the distributed AP, the other distributed AP is searched (S13). After these steps have passed, the distributed AP on the prepared container can actually be tested (S14).

[Deployment of distributed AP for testing]
FIG. 3 is a diagram showing a flow of processing when the distributed AP (AP1) 1410 is deployed in the computer 1000 for the purpose of testing. FIG. 5 is a diagram for supplementarily explaining the flow of this process. Hereinafter, the flow of the deployment process will be described with reference to FIG. 3 while appropriately referring to the flow of deployment in the test mode in FIG. This deployment process is a specific example of the first stage of deployment (S11) and the second stage of deployment (S12) in the description of FIG.

(1) The distributed AP deployment unit 1220 receives AP deployment information 1400 configured from a distributed AP (AP1) 1410 that performs deployment, an execution mode 1420, and automatic conversion rule definition information 1430 of distributed AP runtime information at the time of deployment. Receive as information. Upon receiving the deployment request, the distributed AP deployment unit 1220 determines the execution mode 1420 received as the input information, and starts the distributed AP deployment process (S101).

(2) When a mode other than the actual operation mode (test mode) is designated, a distributed AP information conversion request is sent to the distributed AP runtime information automatic conversion unit 1230 (S102).

(3) The distributed AP execution time information automatic conversion unit 1230 performs automatic conversion rule definition information 1430 and automatic conversion rule common definition information 1300 for the distributed AP (AP1) 1410 to be deployed notified from the distributed AP deployment unit 1220. Conversion of runtime information based on. However, when the automatic conversion rule common definition information 1300 is not defined, the runtime information is converted based on the conversion rule set in advance for each execution mode 1420. Here, the automatic conversion rule definition information 1430 that can be specified for each distributed AP can be omitted, and if specified, it takes precedence over the automatic conversion rule common definition information 1300 (S103).

(4) The distributed AP runtime information automatic conversion unit 1230 notifies the distributed AP deployment unit 1220 that the conversion process has been completed (S104).

(5) The distributed AP deployment unit 1220 deploys the distributed AP (AP1) 1211a in which the runtime information has been converted to the container 1210 (S105).

(6) The distributed AP deployment unit 1220 deploys the test object name (TEST / computer 1 / container / AP1) and object reference information (Reference1a) deployed to the name table access control unit 1241 after deployment to the container 1210. To request storage in the test name table 3213. The test object name uses a designated character string ("TEST" in this embodiment) designated by the automatic conversion rule common definition information 1300 or the automatic conversion rule definition information 1430 (S106).

(7) The name table access control unit 1241 notifies the distributed object base unit 1260 of the received object name (TEST / computer 1 / container / AP1) and object reference information (Reference1a), and stores it in the naming service unit 3210. (S107).

(8) The distributed object base unit 1260 generates communication, and notifies the distributed object base unit 3260 of the computer 3000 of the received object name (TEST / computer 1 / container / AP1) and object reference information (Reference 1a). A request is made to store to the test name table 3213 of the name table 3211 in the naming service unit 3210 (S108).

(9) The distributed object base unit 3260 of the computer 3000 stores the received object name (TEST / computer 1 / container / AP1) and object reference information (in the test name table 3213 in the name table 3211 in the naming service unit 3210). Reference 1a) is stored (S109a). The registration result is as shown in a column 3321 of the test name table registration information 3320.

(10) The naming service unit 3210 notifies the distributed code base unit 3260 of a return code for notifying the end (S110).

(11) The distributed object base unit 3260 generates communication and notifies the distributed object base unit 1260 of the received return code (S111).

(12) The distributed object base unit 1260 notifies the received return code to the name table access control unit 1241 (S112).

(13) The name table access control unit 1241 notifies the distributed AP deployment unit 1220 of a return code notifying the end (S113).

  After the above deployment processing ends, the distributed AP deployment unit 1220 notifies the container 1210 of a return code to notify the termination, and triggered by this, the distributed AP (after conversion of runtime information prepared on the container 1210 ( AP1) 1211a can be started.

[Deployment of production distributed AP]
FIG. 4 is a diagram showing a flow of processing when the distributed AP (AP1) 1410 is deployed in the computer 1000 for production use. Hereinafter, the flow of the deployment process will be described with reference to FIG. 4 while appropriately referring to the flow of the production mode of FIG. The same processes as those in FIG. 3 are denoted by the same reference numerals and description thereof is omitted.

(1) The execution mode is determined in the same manner as in S101 of FIG. 3, and the distributed AP deployment process is started (S101). Note that when the mode during the actual operation is designated or when nothing is designated, a processing request to the distributed AP runtime information automatic conversion unit 1230 is not performed.

The processing from S105 to S108 corresponding to (2) to (5) is the same as that in FIG.

(6) The distributed object base unit 3260 of the computer 3000 stores the received object name (computer 1 / container / AP1) and object reference information (Reference1b) in the production name table 3212 in the name table 3211 in the naming service unit 3210. Is stored. The registration result is as shown in the column 3311 of the production name table registration information 3310 (S109b).

The processing from S110 to S113 corresponding to (7) to (10) is the same as that in FIG.

  After the deployment processing so far, the distributed AP deployment unit 1220 is notified of the return code for notifying the container 1210 of the end, and the distributed application (AP1a) 1211a deployed in the container 1210 receives the received object reference information ( The distributed application (AP2a) 2211a is called using the Reference2a) to request processing.

  As in the case of the test mode, after the above-described deployment processing is completed, the distributed AP deployment unit 1220 notifies the container 1210 of a termination code, and using this as an opportunity, the execution prepared on the container 1210 is executed. It becomes possible to start the distributed AP (AP1) 1211a after the time information conversion.

  FIG. 5 is a diagram illustrating the flow of the deployment process and supplementarily explaining both the test mode and the production mode. The individual steps from S101 to S113 in FIG. 5 have already been described in the description with reference to FIGS. 3 and 4, and the difference between the steps in the test mode and the actual mode will be described here.

  In FIG. 5, steps unique to the test mode are surrounded by a thick frame, and steps performed by the computer managing the name table are surrounded by a double frame. In FIG. 5, S102, S103, S104, and S109a correspond to the steps unique to the test mode. Of these, S102, S103, and S104 have no corresponding steps in the production mode deployment process, and the distributed AP used for the test is converted into runtime information using the automatic conversion rule definition information and the automatic conversion rule common definition information. This is a step in which a distributed AP for testing can be performed.

  S109a corresponds to S109b in the deployment process in the production mode, and this is different in that the test use and the production use are distinguished when stored in the name table.

  These steps, which are different between the test mode and the production mode, particularly S103 and S104, play an important role in the present embodiment in order to form an environment in which the distributed AP in the test mode can be executed in the same environment as the production mode.

[Deployment processing of distributed AP runtime information automatic conversion unit]
FIG. 6 is a diagram illustrating a processing flowchart example of the distributed AP runtime information automatic conversion unit 1230 that is performed when the distributed AP is deployed on the container. This process flowchart example is a diagram collectively showing the flow of the distributed AP deployment process individually described for the test mode and the production mode. The process illustrated in FIG. 6 corresponds to the deployment process (S11) of the distributed AP execution time information automatic conversion unit in FIG.

  First, the execution mode 1420 of the distributed AP to be deployed is confirmed to confirm whether it is a production distributed AP (S301).

  If it is a distributed AP for production (Yes in S301), the process is terminated. If it is a non-production distributed AP (deployment in test mode) (No in S301), either automatic conversion rule common definition information 1330 or automatic conversion rule definition information 1430 for automatically converting runtime information It is confirmed whether one or more are defined (S302).

  If there is no definition (No in S302), the runtime information is converted based on the conversion rule preset in the system for each execution mode 1420, and the process is terminated (S303).

  When either the automatic conversion rule common definition information 1300 or the automatic conversion rule definition information 1430 for automatically converting the runtime information is defined (Yes in S302), the condition of the automatic conversion rule common definition information 1300 And whether or not the runtime information needs to be changed is confirmed (S304).

  If there is a match with the condition (Yes in S304), the runtime information of the distributed AP is changed according to the automatic conversion rule common definition information 1300 (S305).

  If there is no match (No in S304), or after the change of the runtime information in S305, it is confirmed whether the automatic conversion rule definition information 1430 that can be specified for each distributed AP to be deployed is specified ( S306).

  If not specified (No in S306), the process ends. If it is specified (Yes in S306), the conditions of the automatic conversion rule definition information 1430 are confirmed, and it is confirmed whether it is necessary to change the runtime information (S307).

  If there is no match (No at S307), the process is terminated. If there is a match with the conditions (Yes in S307), the runtime information of the distributed AP is changed again according to the automatic conversion rule definition information 1430 (S308), and the process ends.

  The distributed processing of the distributed AP runtime information automatic conversion unit is also applied when the distributed AP to be deployed is a distributed AP of a servlet (see http://java.sun.com/products/servlet/), for example. be able to. In this case, since it is necessary to separate the document root (context root information) for accessing this distributed AP from others for production and testing, a deployment descriptor (deployment descriptor that holds this information before deployment) ) Is changed to the value specified in the automatic conversion rule common definition information 1300 or the automatic conversion rule definition information 1430.

  Also, when the distributed AP to be deployed is a distributed AP of EJB (Enterprise Java Beans, registered trademark, http://java.sun.com/products/ejb/), the distributed AP runtime information automatic conversion unit is also used. Can be deployed. When the distributed AP to be deployed is an EJB (registered trademark) distributed AP, there is a problem in that the distributed AP refers to another EJB (registered trademark) distributed AP. In this case, since it is necessary to separate the reference destination for the production and the test, the deployment descriptor holding this information before deployment is specified by the automatic conversion rule common definition information 1300 or the automatic conversion rule definition information 1430 Change to the specified value.

  As another example of the automatic conversion process, when the execution time information to be converted is the conversion rule and execution mode information in the main storage device used at the time of execution, the conversion at the time of the deployment is not performed. A similar environment can be provided. In this case, the operation can be switched without changing any information such as an AP-specific implementation code or deployment descriptor for testing.

  For example, when there is a definition of “test A” (4500 in FIG. 12) as an example in the distributed AP execution mode 1420 being executed by the distributed AP runtime information automatic conversion unit 1230 when the distributed AP is deployed, the automatic conversion rule The common definition information 1300 and the automatic conversion rule definition information 1430 (for example, “Test A, _testA_, TEST1” (4500) in FIG. 12) are stored as conversion rule and execution mode information in the main storage device used at the time of execution. When a distributed AP is executed after deployment, when it is necessary to separate processing for production and for testing, it is dynamically required by using the conversion rule and execution mode information in the main storage device used at the time of execution. The correct information.

  In the case of an example of referring to another EJB (registered trademark) distributed AP ("computer 2 / container / AP2"), the memory information at the time of execution stored in the conversion unit 1230 at the time of execution of the distributed AP is referred to. When the reference occurs, it is converted into appropriate object reference information (“TEST1 / computer 2 / container / AP2”) and the process is separated.

[Deployment process of name table access control unit]
FIG. 7 is a diagram illustrating a processing flowchart example of the name table access control unit 1241 that is performed when a distributed AP is deployed on a container. FIG. 7 also illustrates the flow of the deployment process by bringing together the test mode and the production mode, not individually, as in FIG. This process corresponds to the deployment process (S12) of the name table access control unit in FIG.

  First, the execution mode 1420 of the distributed AP (AP1) 1410 to be deployed is confirmed to confirm whether it is a production distributed AP (S401).

  If it is a distributed AP for production (Yes in S401), the notified object name and object reference information are registered in the production name table (S402), and the process is terminated.

  If the distributed AP is not for production (deployment in test mode) (No in S401), it is checked whether a test mode name table already exists (S403).

  If it does not exist (No in S403), a test name table is created (S404). If it exists (Yes in S403), the process proceeds to the next step S405.

  If the test mode name table already exists, or after the creation of the test mode name table in S404, register the object name and object reference information notified to the test mode name table ( S405), the process ends.

[Search processing of distributed AP in test mode from distributed AP in test mode]
FIG. 8 shows a case where the distributed AP (AP1a) 1211a refers to the distributed AP (AP2a) 2211a in a state where the distributed AP (AP1a) 1211a and the distributed AP (AP2a) 2211a are both deployed in the test mode on different computers. It is a figure which shows the flow of this object reference information acquisition process. FIG. 10 is a diagram for supplementarily explaining the flow of search processing. Hereinafter, the flow of the distributed AP search process in the test mode will be described with reference to FIG. 8 is a part corresponding to the step of the distributed AP search process (S13) in FIG.

(1) The distributed AP (AP1a) 1211a notifies the name table access control unit 1241 of the object name (TEST / computer 2 / container / AP2) of the distributed AP (AP2a) 2211a and refers to the object of the distributed AP (AP2a) 2211a. Ask for information. Since the object name used for the inquiry has been changed by the runtime information conversion process at the time of distributed AP deployment, it is possible to access the object reference information of the expected execution mode with the acquired object name. Yes (S201a).

(2) The name table access control unit 1241 notifies the distributed object base unit 1260 of the notified object name (TEST / computer 2 / container / AP2), and requests acquisition of object reference information from the name table 3211 ( S202a).

(3) The distributed object base unit 1260 generates communication, notifies the received object name (TEST / computer 2 / container / AP2) to the distributed object base unit 3260 of the computer 3000 and inquires for object reference information (S203). ).

(4) The distributed object platform unit 3260 of the computer 3000 notifies the object reference information by notifying the received object name (TEST / computer 2 / container / AP2) to the name table 3211 in the naming service unit 3210 (S204). .

(5) The naming service unit 3210 searches the reference information of the distributed AP (AP2a) 2211 from the name table 3211 using the received object name (TEST / computer 2 / container / AP2), and distributes the acquired object reference information. The object base unit 3260 is notified. Specifically, the registration information 3322 of the test name table registration information 3320 in the name table registration information 3300 is searched, and the object reference information (Reference 2a) is returned (S205).

(6) The distributed object base unit 3260 generates communication, and notifies the distributed object base unit 1260 of the received search result (Reference 2a) (S206).

(7) The distributed object base unit 1260 notifies the name table access control unit 1241 of the received search result (Reference 2a) (S207).

(8) The name table access control unit 1241 notifies the distributed AP (AP1a) 1211a that is the caller of the received search result (Reference2a) (S208).

  In step S208, the object search process ends, but the distributed AP (AP1a) 1211a then calls the distributed AP (AP2a) 2211a using the received object reference information (Reference2a) to request processing. The execution of the distributed AP is continued. The subsequent movement will be described later in the description of FIG.

  When referring to another distributed AP from the distributed AP deployed on the container 1210 in the above processing, since the runtime information is automatically converted when the distributed AP is deployed, the caller's distributed AP side In this case, the appropriate callee reference information is already held for each execution mode. This makes it possible to make a call request without causing the execution mode to be a problem.

[Search processing of distributed AP in test mode from normal distributed AP]
FIG. 9 shows an example of the flow of acquisition processing when the distributed AP (AP2a) 2211a deployed in the test mode is referenced from a normal (similar to the production) distributed AP (AP1) 1410 on another computer. FIG.

  In the distributed AP (AP1) 1410, when the object name of the reference destination distributed AP (AP2a) 2211a is implemented with the production name (computer 2 / container / AP2), the execution mode 1420 and the argument information at the time of execution are By providing the automatic conversion rule definition information 1430 used when the distributed AP (AP2a) 2211a that is the reference destination is provided, it is possible to perform an operation test of the distributed AP (AP2a) 2211a without changing the implementation of the distributed AP 1410. .

  Hereinafter, the flow of search processing for a distributed AP deployed in a test mode from a normal distributed AP will be described with reference to FIG. 10 as appropriate. 9 is a part corresponding to the step of the distributed AP search process (S13) in FIG. In FIG. 9, since there are many portions of the same processing as in FIG. 8, the same reference numerals as in FIG.

(1) The distributed AP (AP1) 1410 is a parameter information at the time of execution (specifically, a command argument at the time of execution in the case of a distributed AP written in the Java (registered trademark) language, a specification as a system property value, etc.) ) And the automatic conversion rule definition information 1430 used when the distributed AP (AP2a) 2211a that is the reference destination is deployed and the object name (computer 2 / container / AP2) are notified to the name table access control unit 1241. Then, the object reference information of the distributed AP (AP2a) 2211a is inquired (S201b).

(2) The name table access control unit 1241 uses the notified execution mode 1420 and the automatic conversion rule definition information 1430 to notify the object name (computer 2 / container / AP2) that matches the execution mode (TEST). / Computer 2 / container / AP2), notify the distributed object platform 1260 of the changed object name (TEST / computer 2 / container / AP2), and request acquisition of object reference information from the name table ( S202b).

  Hereinafter, S203 to S208 corresponding to (3) to (8) are the same as those in FIG. Similar to the search process from the distributed AP deployed in the test mode of FIG. 8, the process of executing the distributed AP continues following the search process from the normal distributed AP shown in FIG. This process will be described later.

  In the search process, only the first two steps are different between the case of the distributed AP deployed in the test mode and the case of the normal distributed AP as shown in FIG. In FIG. 10, when the two processes are different, the steps for processing from the distributed AP deployed in the test mode are shown surrounded by a thick frame. In the first two steps, the difference between the two operations is that the search processing from the distributed AP deployed in the test mode is different from the normal distributed AP in advance. This is because since necessary information is given, there is no need to change the object name based on the execution mode or automatic conversion rule definition information specified by the argument information.

  FIG. 11 is a diagram for explaining a distributed AP call process performed subsequent to the deployment process and the search process. The steps shown in FIG. 11 are the same as the steps S209 to S218 performed subsequent to the search processing in S201 to S208 shown in FIGS. At this stage, there is no difference depending on whether or not the test mode is deployed or searched, and the distributed AP call processing performed after each of the deployment processing and search processing described so far is illustrated in FIG. 11 is processed in the same manner as shown in FIG.

  After the end of the search process, first, the object name and the acquired reference information are notified, and the distributed AP (AP2a) 2211a is called to the remote call unit 1250 (S209). Next, the remote calling unit 1250 notifies the distributed object base unit 1260 of the object name and the object reference information (S210). Upon receiving the object name and the object reference information, the distributed object base unit 1260 generates communication and notifies the object name and the object reference information to the distributed object base unit of the computer on which the target object to be referred is deployed (S211).

  The distributed object base unit of the computer that has received the notified target object that has received the notification notifies the remote calling unit of the object reference information and requests the calling of the referenced target (S212). The remote calling unit starts by calling the distributed AP to be referenced that is deployed in the container (S213). After the distributed AP ends, the distributed AP notifies the remote call unit of a return code indicating the end (S214). The remote calling unit notifies the distributed object base unit of a return code indicating that the called distributed AP has ended (S215). Then, the distributed object base unit generates communication and notifies the return code to the calling computer (S216).

  The distributed object base unit of the calling computer notifies the return code to the remote calling unit (S217). The remote calling unit notifies the calling application of the distributed application that the called application has ended (S218).

[Automatic conversion rule common definition information]
FIG. 12 is a data structure diagram showing an example of the automatic conversion rule common definition information 1300 according to an embodiment of the present invention. In this embodiment, a distributed AP such as a servlet or EJB (registered trademark) is assumed as a distributed AP, and a context route identifier is used as a conversion rule in a form that is paired with an execution mode 4100 that indicates the type of deployment / execution of the distributed AP. 2200 and name table route identifier 4300 are defined. This automatic conversion rule common definition information is referred to when the distributed AP is deployed. The context root identifier 4200 is used when the distributed AP is a servlet, and automatically changes the information of the document root (context root) for accessing the servlet from others depending on the execution mode. The name table route identifier 4300 is used when the object reference information of the distributed AP is registered in the name table 3211 and the route information of the registered object name is automatically changed according to the execution mode.

  As a specific example of execution using this data structure, for example, the deployment of a Web distributed AP having the information that the context root is “jsptobean /” and the object name is “WebAP1” is called execution mode-production mode (4400). In the execution mode-test A mode (4500), the context root is “jsptobean /” in the production mode and “_testA_ / jsptobean /” in the test A mode. Also, the object name registered in the name table 3211 is “WebAP1” in the production mode and “TEST1 / WebAP1” in the test A mode. FIG. 13 is a diagram for explaining automatic conversion of the deployment descriptor, and shows an example before and after converting the deployment descriptor in the test A mode.

[Automatic conversion rule definition information]
FIG. 14 is a data structure diagram showing an example of automatic conversion rule definition information 1430 according to an embodiment of the present invention. In an example of the present embodiment, two conversion rules, a context route identifier 4200 and a name table route identifier 4300, are defined. This automatic conversion rule definition information is also referred to when a distributed AP is deployed or a distributed application is executed.

  As described so far, in this embodiment, by providing a method for converting the runtime information of the test distributed AP, etc., the same computer that is executing the production distributed AP can use the production distributed AP. Tests can be executed correctly without making incorrect references to APs or interfering with distributed APs in production. This makes it possible to perform a highly reliable test on a distributed AP, which has been difficult in the past, on a computer on which the actual distributed AP is actually running, and easily, after the test has been completed, the distributed AP can be easily performed. Can be run on the same computer. Furthermore, the method employed in the present embodiment is a method that can be applied to general distributed APs including distributed web APs, regardless of the development language and development environment. Thereby, it is possible to realize a test of a distributed AP with low cost and high reliability.

  This test method is executed by a computer having a CPU and a main storage device, and a program for realizing the test method can be stored in an information storage medium, and the program is read into the main storage device of the computer. The computer can be a distributed application test apparatus equipped with the test method of the present invention.

It is a figure which shows one Embodiment of the distributed object system of this invention. It is a figure which shows the flow of a process until it performs test execution of distributed AP. It is a figure which shows the flow of a process at the time of deploying distributed AP to a computer for the purpose of a test. It is a figure which shows the flow of a process at the time of deploying distributed AP to a computer for production. It is a figure for supplementarily explaining the flow of a deployment process. It is a figure which shows the example of a process flowchart of a distributed AP execution time information automatic conversion part. It is a figure which shows the example of a process flowchart of a name table access control part. It is a figure which shows the flow of the object reference information acquisition process in the case where two distributed APs are deployed in a test mode on different computers, and one distributed AP is referred to another distributed AP. It is a figure which shows an example of the flow of an acquisition process in the case of referring the distributed AP deployed by the test mode from the normal distributed AP on another computer. It is a figure explaining supplementarily the flow of the reference information search process from the distributed AP deployed in the test mode and the normal distributed AP. It is a figure explaining the process of calling of distributed AP. It is a data structure figure which shows an example of automatic conversion rule common definition information. It is a figure explaining the automatic conversion of a deployment descriptor. It is a data structure figure which shows an example of automatic conversion rule definition information.

Explanation of symbols

100 network 1000 computer 1
2000 Calculator 2
3000 Calculator 3
1100, 2100, 3100 CPU
1200, 2200, 3200 Main storage device 1210, 2210 Container 1211, 1211a, 2211b, 1410 Distributed application (distributed AP, object)
1220, 2220 distributed application deployment unit (distributed AP deployment unit)
1230, 2230 Distributed application runtime information automatic conversion unit (Distributed AP runtime information automatic conversion unit)
1240, 2240 Naming interface unit 1241, 2241 Name table access control unit 1250, 2250 Remote call unit 1260, 2260, 3260 Distributed object base unit 1300 Automatic conversion rule common definition information 1400 Application deployment information (AP deployment information)
1420 Execution Mode 1430 Automatic Conversion Rule Definition Information 3210 Naming Service Section 3211 Name Table 3212 Production Name Table 3213 Test Name Table

Claims (6)

A method for testing a distributed application in a distributed application execution environment for a computer system including one or more computers to execute the distributed application,
The computer system is
Determines the execution mode that indicates whether the distributed application is executed for production or testing, and automatically converts the information used when executing the distributed application and the distributed application deployment unit that deploys the distributed application to the execution environment A distributed application runtime information automatic conversion unit that converts according to a rule, a naming service unit that manages object name reference information when referring to a distributed object name assigned to a distributed application called by the distributed application, and a distributed object of the naming service unit A naming interface unit to be referred to, a distributed object base unit that generates communication between computers, a distributed application, and a container that provides an environment for executing the distributed application.
In the computer system,
The distributed application deployment unit adds the execution mode in units of distributed applications and deploys distributed applications.
The distributed application runtime information automatic conversion unit converts the runtime information of the distributed application to be executed when the execution mode is the test mode,
The naming service unit individually manages the object name and object reference information of the distributed application sent via the communication by the distributed object base unit by the naming interface unit in different areas for each execution mode,
The container stores the distributed application;
A distributed application test method, wherein the computer system simultaneously executes a plurality of distributed applications having the same object name. The distributed application test method according to claim 1,
The computer system is
A remote call unit that calls the distributed object when the distributed object is deployed on a computer different from a distributed application that calls the distributed object;
Based on the object reference information referenced by the naming interface unit,
By calling the distributed application deployed on different computers by the remote calling unit via communication by the distributed object platform unit,
A distributed application test method, wherein the computer system simultaneously executes a plurality of distributed applications having the same object name. The distributed application test method according to claim 2,
In a distributed application deployed with the execution mode added in units of distributed applications,
The distributed application runtime information automatic conversion unit,
Runtime information of distributed application consisting of deployment code, deployment identifier including execution environment, or conversion rules and execution mode information used at the time of execution,
When the execution mode is specified as test mode, automatic conversion or automatic setting is performed.
The distributed application test method, wherein the automatic conversion and the automatic setting are not performed when the production mode is designated. The distributed application test method according to claim 3,
In automatic conversion according to the execution mode performed by the distributed application runtime information automatic conversion unit,
The automatic conversion rule definition information is
Includes automatic conversion rule definition information common to computer systems defined in association with execution modes, and individual automatic conversion rule definition information defined for each distributed application to be deployed.
When automatic conversion rule definition information describing an object name conversion rule is specified in the distributed application reference information, the distributed application runtime information automatic conversion unit performs the runtime information according to the content of the automatic conversion rule definition information. Conversion
A distributed application test method, wherein when automatic conversion rule definition information is not specified, a distributed application runtime information automatic conversion unit performs predetermined conversion to an execution mode. The distributed application test method according to claim 4, wherein when the distributed application in the test mode includes a web distributed application, the web distributed application is converted at the time of deployment or execution.
When applying the automatic conversion rule definition information when deploying a web distributed application,
The URL information of the context root, which is a place for storing resources necessary for the web distributed application when accessed from the web browser, is converted after executing the runtime information using the automatic conversion rule definition information and then deployed. ,
When applying the automatic conversion rule definition information when executing the web distributed application,
The automatic conversion rule definition information and the execution mode information are stored in the main memory without performing conversion at the time of deployment, and sequentially applied to the web distributed application according to the execution mode information stored in the main memory by the container. A distributed application testing method characterized by performing the necessary transformations. 6. A distributed application test apparatus, wherein a computer system comprising one or more computers having a main storage device and a central processing unit executes the test method according to any one of claims 1 to 5. .

Images