JP5653112B2 - Command execution result abnormality detection system and method - Google Patents

Description

  The present invention relates to a command execution result abnormality detection system and method, and more particularly, in a system configured with a plurality of servers, when a command is executed collectively by each server, the abnormality of the command execution result is detected. The present invention relates to a command execution result abnormality detection system and method that can detect presence or absence.

  In a system configured with multiple servers, various technologies as described below are available as conventional technologies that can automate the execution and monitoring of processes and reduce server operation man-hours and improve operational accuracy. The technology is known.

  That is, the prior art described in Patent Document 1 proposes a technique related to a server group display monitoring device that can simultaneously and reliably check the operation status of a plurality of servers and whether or not a failure has occurred on one display device. .

  In addition, the prior art described in Patent Document 2 is a multiple server that allows a user to log in to all systems by simply entering a single user ID and password when using multiple systems simultaneously. A technology related to an inter-login linkage system, a client device, a login management device, a server device, and a storage medium is proposed.

  In addition, the prior art described in Patent Document 3 improves the efficiency of operation management by performing the same level of management as when an operation manager is assigned to each server, and reducing the load on the network and management server. We have proposed a technology that can greatly improve reliability.

  The prior art described in Patent Document 4 provides a function for centrally managing job flows within and between servers as a system using a common means from a multi-vendor system composed of a plurality of servers. It proposes a technology related to a method for reducing the burden of failure recovery processing and system management in the event of a failure.

  The prior art described in Patent Document 5 proposes a technology related to a highly reliable server system that can check all data processed by a server and diagnose a failure.

Japanese Patent Laid-Open No. 11-175180 JP 2002-183094 A JP 2000-250833 A JP 2002-32242 A JP 2001-175545 A

  In recent years, systems that operate a plurality of servers of the same application in parallel are increasing due to the ease of building a distributed system due to a decrease in unit price and the development of new technologies such as clusters and clouds.

  However, when work is performed using a system composed of a plurality of servers of the same purpose as described above, the same work must be performed for each server constituting the system. However, the larger the system, the more man-hours will be required. Moreover, the operation is redundant and wasteful, and can be a hotbed for human errors such as typographical errors. For this reason, various automation measures are being promoted, but an increase in man-hours cannot be avoided.

  In general, when executing the same command for multiple servers, the operator logs in to each server, causes the command to be executed for each server, checks whether there are any abnormalities in the execution results of each server, and It is necessary to compare whether there is a difference in the results of. For this reason, when the number of servers increases, it becomes very difficult to perform a visual check.

  In addition, a simple means of confirming the end code is often used as automation of abnormality detection. However, this method cannot detect an unintended normal end.

  In addition, there is a method of creating a command list used for work in advance, but this method cannot be handled by simple copy and paste when parameters are different for each server.

  There is also a danger that the system will be destroyed if the work is performed without being verified when the work is performed.

  And technology to execute commands in cooperation with multiple servers, technology to execute commands in a batch, technology to automatically execute commands according to predetermined procedures, judgment of abnormalities during execution and execution results Techniques such as techniques to be performed have been proposed by the above-described conventional techniques, but these conventional techniques have the following problems.

  The prior art described in Patent Document 1 is to check a failure by dividing one screen and displaying the screen of a server to be worked on. However, the number of servers increases as the number of servers to be processed increases. Therefore, it is necessary to divide the screen only, and there is a possibility that the image cannot be displayed. In addition, since this prior art assumes that the operation is performed on a server in which a failure has occurred, only the operation of a single server can be performed, and operations can be performed on multiple servers at once. It has the problem that it cannot be done. Furthermore, in this conventional technology, it is necessary to connect three cables per processing server, and if there are 50 target servers, a total of 150 cables need to be physically connected. In addition, since a special device is required, there is a problem that an extra cost is increased.

  The conventional technique described in Patent Document 2 is like so-called single sign-on, and can perform authentication for a plurality of servers by one login. Since it is only the part related to login that can reduce the work man-hours in the batch operation for servers, after all, it is necessary to operate the actual work for each server, and it can not be expected to significantly reduce man-hours. In addition, a separate login management device is required, leading to an extra cost increase.

  The prior art described in Patent Document 3 monitors log data output at the time of command execution in the server, and transmits the information to the management server when log data that matches the condition defined in advance is output. Filtering and automatic log transfer. This prior art can output a command execution result to a log to detect an abnormality, but it is necessary to define the definition contents in advance. For this reason, this prior art cannot compare the difference with the log data of other servers and can only deal with a known failure. The problem is that it cannot be judged. Although this prior art can output and monitor the command execution result in a log, it does not have a function to execute the command, so each server must perform an operation for executing the command. Has the problem.

  The prior art described in Patent Document 4 is to perform centralized management of complicated processing, branching, and occurrence of abnormalities by linking jobs registered commands in a multi-vendor environment between servers. However, it is possible to execute commands in the form of jobs, but conversely, in order to execute the same command on servers of the same purpose, it is necessary to register jobs for that number of machines, which is very laborious It has the problem of becoming larger. In addition, because this conventional technology also depends on the exit code of the command registered in the job for abnormal determination of the job execution result, it cannot be compared with the execution result of other servers, and it terminates normally unintentionally. In such a case, there is a problem that it is impossible to detect the case where the execution result is different for only a specific server.

  The prior art described in Patent Document 5 monitors all processes executed by a plurality of servers that perform the same operation synchronously and compares each other's data to detect an abnormality. Since it is necessary to newly construct a hardware mechanism inside, there is a problem that it is very difficult to apply this technology to an existing server in operation. In addition, this conventional technique simply compares the difference in data contents flowing on the host bus and cannot determine the cause. If the cause is investigated from the data on the host bus. To implement, there is a problem that it is necessary to secure personnel who are familiar with the data flowing on the host bus. In addition, this conventional technique has a monitoring function for determining a difference in processing in which a plurality of servers are linked, but does not have means for executing the processing.

  The object of the present invention is to solve the problems of the prior art as described above, and absorb different parameters for each server included in the command using a unique variable for a plurality of servers of the same use, and collect them collectively. Provide a command execution result abnormality detection system and method that enable command execution, compare execution results for each server, determine differences, and automatically determine whether the execution result is abnormal or normal There is to do.

  According to the present invention, the object is to execute a business process by causing each server of a server group composed of a plurality of servers to execute a command, determine a difference between execution results of each server, and In the anomaly detection system for command execution results to detect anomalies, the server group registration database, the process registration database, the variable substitution registration database, and the server name from the server group registration database as a server constituting the server group And the verification flag, and the target server information definition section that is defined as the verification server and the production server that is the execution target server of the server business process for which the verification flag is set, and the business process name from the process registration database Load the corresponding command and execute it as a business process execution target An execution command information definition section that defines commands and a server name and replacement character string are read from the variable substitution registration database, and the unique variable included in the execution target command defined by the execution command information definition section is associated with each server. A variable substitution processing unit that performs substitution processing on the replacement character string to be performed, a verification result confirmation unit that determines whether the execution result of the command verified by the verification server is normal or abnormal, and the production server This is achieved by providing a difference processing unit that compares the execution results of the servers in the command execution results to detect a difference and discriminates between normal end and abnormal end.

  According to the present invention, batch execution is possible when individual servers constituting a server group perform the same work, and the work man-hour is reduced to the work man-hour for one server, no matter how large the system is. It is possible to reduce human error intervening points.

1 is a block diagram showing a schematic configuration of a system to which the present invention including a command execution result abnormality detection system according to an embodiment of the present invention is applied. It is a flowchart explaining the outline | summary of the processing operation in an execution part. It is a flowchart explaining the detail of the processing operation in a target server information definition part. It is a flowchart explaining the detail of the processing operation in an execution command information definition part. It is a flowchart explaining the detail of the processing operation in a variable substitution process part. It is a flowchart explaining the detail of the processing operation in a verification result confirmation part. It is a flowchart explaining the detail of the processing operation in a difference process part. It is a flowchart explaining the processing operation in a database registration process part. It is a figure which shows the item description and data example explaining the structure of a business registration database. It is a figure which shows the item description and data example explaining the structure of a process registration database. It is a figure which shows the item description and data example explaining the structure of a server group registration database. It is a figure which shows the item description and data example explaining the structure of a server registration database. It is a figure which shows the item description and data example explaining the structure of a variable substitution registration database.

  Embodiments of a command execution result abnormality detection system and method according to the present invention will be described below in detail with reference to the drawings.

  FIG. 1 is a block diagram showing a schematic configuration of a system to which the present invention is applied, including a command execution result abnormality detection system according to an embodiment of the present invention.

  As shown in FIG. 1, a system to which the present invention is applied includes a client machine 10 operated by a worker, and a server group 12 including a plurality of servers that execute commands transmitted from the client machine 10. Are connected to each other via a network 11.

  The command executed in the embodiment of the present invention is executed based on a command using key input by Telnet, but the command is executed by another key input. It is applicable if it is a system using.

  Although not shown, the client machine 10 includes a CPU, a main memory, a storage device such as an HDD, a keyboard used by workers, an input device such as a mouse, a display device, and the like. In the main memory, programs for constructing various processing units provided for the present invention to be described later are loaded from a storage device and stored. These programs are executed by the CPU to construct processing units having respective functions.

  The client machine 10 includes a command execution result abnormality detection system (hereinafter simply referred to as an abnormality detection system) 100 operated by a worker.

  The abnormality detection system 100 includes a registration unit 110 that performs an operation on a database stored in the client machine 10 and an execution unit 120 that is a process necessary for the server group 12 to execute a command. The database stored in the client machine 10 includes a business registration database 112, a process registration database 113, a server group registration database 114, a server registration database 115, and a variable replacement registration database 116.

  The registration unit 110 includes a database registration processing unit 111. The database registration processing unit 111 includes the above-described business registration database 112, process registration database 113, server group registration database 114, server registration database 115, and variable substitution registration database 116. For each database, new data registration, registration content change, and registration content deletion are performed.

  Next, the configuration of each of the databases described above will be described with reference to FIGS.

  FIG. 9 is a diagram showing item descriptions and data examples for explaining the configuration of the business registration database 112.

  The business registration database 112 is a database that records information related to a business for executing a command by using the present invention performed by a worker. As shown in the data example, a business name that is an identifier of a business that uses the present invention. And a process name that is an identifier of a process associated with the business and a server group name that is an identifier of a server group that is the execution target of the business.

  As described in the item description, the business name is a business name using the system of the present invention, and is determined by directly inputting a character string by a worker at the time of new registration. Further, as described above, the process name is the name of the process associated with the business, and is selected from information registered in advance in another database at the time of new registration. Since the process name recorded in the business registration database 112 is provided with necessary information from the process registration database 113, the process name needs to be recorded in advance in the process registration database 113.

  In addition, as described above, the server group name recorded in the business registration database 112 is the name of the server group that is the target of the business, and is selected from information registered in advance in another database at the time of new registration. Is done. The server group name recorded in the business registration database 112 is provided with necessary information from the server group registration database 114. Therefore, the server group name needs to be registered in the server group registration database 114 in advance. There is.

  FIG. 10 is a diagram showing item descriptions and data examples for explaining the configuration of the process registration database 113.

  The process registration database 113 is a database that records information related to the contents of the process linked to the work recorded in the work registration database 112. As shown in the data example, the process registration database 113 is tied to the work using the present invention. The process name is an identifier of a process to be attached, a command executed by the server group 12 associated with the process, and a line number indicating the execution order of the command.

  As described in the item description, the process name is a process name in which the system of the present invention is used, and is determined by directly inputting a character string by a worker at the time of new registration. The line number is a number indicating the execution order in this process, and is determined by directly inputting a character string by a worker at the time of new registration. The command is a command executed in the processing here, and is determined by directly inputting a character string by a worker at the time of new registration.

  In the above description, the process name recorded in the process registration database 113 is referred to from the outside by the process name of the work registration database 112, and the command and line number associated with the work from the process name of the work registration database 112 are described. And can be derived.

  FIG. 11 is a diagram showing item descriptions and data examples for explaining the configuration of the server group registration database 114.

  The server group registration database 114 is a database that records information related to server groups that are targets of operations performed by workers. As shown in the data example, the servers that are targets of execution of operations using the present invention are grouped. The server group name that is the identifier of the server group that enables registration of multiple servers in one group, the server name that is the identifier of the server linked to the server group, and the server executing the business In FIG. 4, a verification flag indicating whether or not the verification execution is performed (the verification target is “1” and the non-verification target is “0”). In the present invention, the number of servers in which the verification flag is set in the server group is limited to one, and it is not permitted to set a plurality of verification flags.

  As described in the item description, the server group name recorded in the server group registration database 114 is a server group name using the system of the present invention, and a character string is directly input by a worker at the time of new registration. Determined. The server name is a server name of a server included in the server group described above, and is selected from information registered in advance in another database at the time of new registration. The verification flag is selected from information registered in advance in another database at the time of new registration.

  In the above description, the server group name is externally referred to by the server group name of the business registration database 112, and the server name and verification flag associated with the business are derived from the server group name of the business registration database 112. Can do.

  The server name recorded in the server group registration database 114 is provided with necessary information from the server registration database 115. Therefore, the server name needs to be recorded in the server registration database 115 in advance.

  FIG. 12 is a diagram showing item descriptions and data examples for explaining the configuration of the server registration database 115.

  The server registration database 115 is a database that records information related to a server linked to a server group that is a target of work performed by a worker. The server name that is the identifier of the server linked to the server group, the address indicating the location of the server on the network, the user required to log in to the server, and the password defined for the user It is comprised by. As described in the item description, all of these items are determined by directly inputting a character string by an operator at the time of new registration.

  In the above description, the server name recorded in the server registration database 115 is referred to from the outside by the server name in the server group registration database 114, and is associated with the server group by the server name in the server group registration database 114. The server address, user and password can be derived.

  FIG. 13 is a diagram showing item descriptions and data examples for explaining the configuration of the variable replacement registration database 116.

  The variable replacement registration database 116 is necessary for replacing a unique variable that is a variable uniquely used in the present invention in a command transmitted to execute a command in the server group 12. As shown in the data example, a unique variable that is a variable character string to be replaced and included in a command executed in a task using the present invention is replaced with the unique variable. Server name to be processed, and a replacement character string that is a character string after replacement of the unique variable. As described in the item description, the items of the original variable and the replacement character string are determined by inputting the character string directly by the worker at the time of new registration, and the server name is stored in advance in another database at the time of new registration. Is selected from the information registered in.

  In the above description, the server name recorded in the variable replacement registration database 116 is provided with necessary information from the server registration database 115. Therefore, the server name needs to be recorded in the server registration database 115 in advance. .

  Returning to the reference in FIG. 1, the execution unit 120 includes a job selection screen 121 that is a screen for selecting a job to be executed by a worker, a target server information definition unit 122 that defines a server that executes the job, and the job. An execution command information definition unit 123 that defines a command to be executed in step 1, a variable substitution processing unit 124 that processes a unique variable included in the command to be executed, and a target server defined by the target server information definition unit 122 described above. The command transmission 125 for transmitting the execution command defined by the execution command information definition unit 123 and having the unique variable replaced by the variable replacement processing unit 124 and the execution result of the command executed by the server group 12 are received. Of the execution results received by the result reception 126 and the execution result reception 126, the verification execution result indicates normal termination and abnormal termination. It is composed of and the verification result check unit 127 discriminates detects the difference in the verification execution results in no execution result and the difference processing unit 128 for discriminating successful and unsuccessful.

  Next, the processing operation in each processing unit shown in FIG. 1 will be described with reference to the flowcharts shown in FIGS.

  FIG. 8 is a flowchart for explaining the processing operation in the database registration processing unit 111. First, the processing operation in the database registration processing unit 111 will be described with reference to FIG.

(1) First, the worker uses the client machine 10 to display a registration target type selection screen by the system of the present invention, and selects one type of item to be registered on the screen. There are five types of items that can be selected: process, command, service, server, and substitution variable (step 800).

(2) The type of the item selected by the worker on the above-described screen is stored in the system as the registration target type. Then, the registration content selection screen is displayed by the system of the present invention. The worker uses the registration content selection screen to select one type of registration to be performed on the registration target type selected above. There are three items of contents that can be selected: new registration, registration change, and registration deletion (steps 801 and 802).

(3) It is determined whether the registered content as the work selected by the worker on the registered content selection screen displayed in the processing of step 802 is new registration, registration change, or registration deletion (step 803).

(4) If it is determined in step 803 that the registered content selected by the worker is a new registration, a new registration screen is displayed by the system of the present invention. In this new registration screen, based on the registration target type set in the processing of step 801, the input items necessary for newly registering the target type are determined from the target database column, and the necessary items are displayed to the worker. Information for input is displayed (step 810).

(5) When the worker inputs necessary items without omission according to the new registration screen displayed in the processing of step 810, the system of the present invention confirms whether all necessary items input by the worker are correct. Execute the process. This check process is a process to check the consistency with the database, and check the numerical value for the character string item, the character string for the numerical value item, the number of digits, the illegal character use, and the blank item restriction. Performs reconfirmation of the input items by visual inspection (steps 811 and 812).

(6) If incompleteness is confirmed in the registered contents in the confirmation process in step 812, or if correction of the input contents is selected by the worker, the process returns to the process from step 811 for inputting the necessary items. When the process is repeated and no incompleteness is confirmed in the registered contents, a process of updating the database based on the input item is executed. This update process will be described later (step 804).

(7) If it is determined in step 803 that the registered content selected by the worker is a registration change, the process proceeds to a change target selection process, and the registration target type set in the process of step 801 is set by the system of the present invention. Based on this, the names of candidates for registration change registered in the target type of the database are read from the database and displayed in a list. The worker selects a registration change target from the displayed change candidate list and causes the system to hold the selected target (steps 820 and 821).

(8) Next, a registration change screen display process is performed in which the items read from the database in step 821 and held in the system are displayed on the screen in a state that can be changed by the worker (step 822).

(9) When the worker changes the changeable items as necessary according to the registration change screen displayed in the process of step 822, the system according to the present invention allows the operator to change and input the input items. A process for confirming whether or not everything is correct is executed. This check process is a process to check the consistency with the database, and check the numerical value for the character string item, the character string for the numerical value item, the number of digits exceeded, illegal character use, and blank item restriction. The worker reconfirms the input items by visual inspection (steps 823 and 824).

(10) If the registration content is confirmed to be incomplete in the confirmation process in step 824, or if correction of the input content is selected by the worker, the process returns to the process from step 823 for inputting the change item. When the process is repeated and no incompleteness is confirmed in the registered contents, a process of updating the database based on the input item is executed. This update process will be described later (step 804).

(12) If it is determined in step 803 that the registered content selected by the worker is deletion of registration, the process proceeds to deletion target selection processing, and the registration target type set in step 801 by the system of the present invention is set. Based on this, names that are candidates for registration deletion targets registered in the target type of the database are read from the database and displayed in a list. The worker selects a registration deletion target from the displayed deletion candidate list, and causes the system to hold the selected target. Then, the item associated with the target held in the system is read from the corresponding database and held in the system (steps 830 and 831).

(13) Next, a registration deletion screen display process for displaying the items read from the database in the process of step 831 and held in the system on the screen for visual confirmation by the worker is performed (step 832).

(14) The worker performs a deletion target confirmation process for visually confirming whether the display item displayed on the registration deletion screen displayed in the process of Step 832 is a deletion target. In this confirmation process, if an error is confirmed by the worker in the deletion target, the process returns to the process from step 830 for selecting the deletion target, and the process is repeated. If no error is confirmed in the deletion target, it is input. Execute the process to update the database based on the item. This update process will be described later (steps 833 and 804).

  Next, the relevant database update process in step 804 executed after the confirmation process in steps 812, 824, and 833 described above will be described. This database update process 804 is a process executed when there is no problem in the confirmation of the registration content confirmation 812, the change content confirmation 824, and the deletion target confirmation 833.

  If the database update process in step 804 is a registration content confirmation process, the registration content confirmation process is performed on the database based on the registration target type set in the process in step 801. This is a process for newly registering the registered contents confirmed in step (b).

  In the process of updating the corresponding database in step 804, if the process in the previous step is a change content confirmation process, the change content confirmation process is performed on the database based on the registration target type set in the process in step 801. This is a process of reflecting the changed content confirmed in the step 820 with respect to the object selected in the registration object selection process in step 820.

  In the process of updating the corresponding database in step 804, if the process in the previous step is a deletion target confirmation process, the deletion target confirmation in step 833 in the database based on the registration target type set in the process in step 801 is performed. This is a process for deleting the information to be deleted confirmed in the process.

  By the database update by the process of step 804 described above, a series of processes in the flowchart of the database registration processing unit described with reference to FIG. 8 is completed, and the process returns to the process from step 800 for displaying the registration target type selection screen and registered on that screen. Processing can be performed by allowing the operator to select one of the other types of items to be processed.

  FIG. 2 is a flowchart for explaining the outline of the processing operation in the execution unit 120. Next, the processing operation in the execution unit 120 will be described with reference to FIG.

(1) First, the worker uses the client machine 20 to display a job selection screen using the system of the present invention. When displaying the business selection screen, the system of the present invention reads business names that are candidates for business selection from the business registration database 220 and displays them in a list on the screen. The worker selects one business name to be executed from the displayed list of screens. The process name and server group name, which are items associated with the business name selected by the worker, are read from the business registration database 220 and held in the system (step 200).

(2) Next, when a command is executed in the server group 21, a command pointer indicating a command position being executed among a plurality of commands is set to an initial position (line number 1). This command pointer can realize complicated processing including verification execution, verification result judgment, actual execution, execution result comparison, and abnormality confirmation process for each command execution in a plurality of servers without losing sight of the command execution position. Yes (step 201).

(3) Thereafter, the target server information definition process at step 202, the execution command information definition process at step 203, the variable substitution process at step 204, the command transmission process at step 205, and the command by the server group 12 at step 206 The execution process and the execution result reception process from the server group 12 in step 207 are sequentially executed, and further, the difference process in step 209 or the verification result confirmation process in step 210 is executed. Then, after the difference process in step 209, the process proceeds to the process of step 301 described later by the flow shown in FIG. 3, and after the verification result confirmation process in step 210, the process of step 304 described later by the flow shown in FIG. Migrate to These processes will be described later in order.

  FIG. 3 is a flowchart for explaining the details of the processing operation in the target server information definition unit 122, which will be described next. The processing operation in the target server information definition unit described here is a target server information definition process for defining a target server to be a business execution target in step 202 in the flow shown in FIG. Note that the input in this processing is the server group registration database 221, and the output is the server list 320 (230 in FIG. 2), which is a list of all servers to be executed in the business, and the flow of FIG. A target server information definition 321 (231 in FIG. 2) that defines a server to be executed in a loop.

(1) The processing is shifted from the command pointer initialization process in step 201 of the flow shown in FIG. 2 to start the target server information definition process. First, as shown in FIG. The item linked to the server group name in the job selection screen held in the system in the process at step 200 of the flow described is read from the server group registration database 310. The items read here are a server name and a verification flag. The server name and the verification flag that are read items are held in the system as a server list 320 (steps 30 and 300).

(2) Next, it is confirmed from the server list 320 held in the process of step 300 whether or not there is a server for which the verification flag is set, and the existence of the server for which the verification flag is set is confirmed. In this case, only servers for which the verification flag is set in the server list 320 are set as execution targets, and the target server information definition 321 is stored in the system. The execution target server held here is called the verification server, and the verification server executes the verification, so that the execution result can be grasped before executing on other than the verification server. (Steps 301 and 302).

(3) If it is confirmed in step 301 that the existence of a server for which the verification flag is set in the server list 320 is not confirmed, all the servers set in the server list 320 are set as execution targets, and target server information The definition 321 is retained in the system. In this case, since the execution result of the command cannot be grasped in advance, the abnormality of the command itself cannot be detected and the failure influence range cannot be narrowed (step 303).

(4) The transition after the difference processing at step 209 in the flow shown in FIG. 2 is performed from the first connection 305 and shifts to the processing at step 301, and after the verification result confirmation processing at step 210. The transition is performed from the second connection 306, and the server to be executed is set as a target server information definition 321 in the system except for the verification server. As a result, when the verification execution result by the verification server is normal, it is possible to safely execute the command with an object other than the verification server as an execution target (step 304).

(5) After the execution target is stored in the system as the target server information definition 321 in the processing of step 302, 303, or 304, the processing of the target server information definition, which is the processing here, is terminated and illustrated in FIG. The processing is returned to the flow (step 31).

  FIG. 4 is a flowchart for explaining the details of the processing operation in the execution command information definition unit 123. Next, this will be described. The processing operation in the execution command information definition unit 123 described here is an execution command information definition process for defining a command to be executed in step 203 in the flow shown in FIG. Note that the input in this process is the process registration database 113, and the output is a command list 420 (232 in FIG. 2) which is a list of all commands to be executed in the job, and the flow loop of FIG. Is an execution command information definition (233 in FIG. 2) defining a command to be executed.

(1) When the process transitions from the target server information definition process in step 202 of the flow described with reference to FIG. 2, the execution command information definition process starts here. First, as shown in FIG. The item associated with the process name on the job selection screen held in the system in the process at Step 200 of the flow described is read from the process registration database 410. The items read here are line numbers and commands. The line numbers and commands that are read items are held in the system as a command list 420 (steps 40 and 400).

(2) Next, referring to the command point held in the system, the command point and the line number held in the command list 420 are compared to find a matching item, and the command with the matching line number is searched. Take out (step 401).

(3) The command extracted in the process of step 401 is held in the system as the execution command information definition 421. Thereafter, the process of the execution command information definition, which is the process here, is terminated, and the flow described with reference to FIG. The process is returned to (Steps 402 and 41).

  FIG. 5 is a flowchart for explaining the details of the processing operation in the variable substitution processing unit 124, which will be described next. The processing operation in the variable replacement processing unit 124 described here is a variable replacement that performs special processing when a unique variable is included in the command to be executed in step 204 in the flow shown in FIG. It is processing. Note that the input in this processing is the variable replacement registration database 511 (223 in FIG. 2), and the output is a replacement list 520 (234 in FIG. 2) referred to when replacing the unique variable.

(1) When the process is shifted from the execution command information definition process in step 203 of the flow shown in FIG. 2, the variable substitution process is started here. First, the steps of the flow shown in FIG. The execution command information definition 233 held in the system in the execution command information definition processing at 203 is referred to and it is confirmed whether or not the command includes an independent variable (steps 50 and 500).

(2) If it is confirmed in step 500 that the command includes an unique variable, the item linked to the unique variable is read from the variable replacement registration database 510. The items read here are the server name and the replacement character string. The server name and replacement character string that are read items are stored in the system as a replacement list 520. If a plurality of unique variables are included, a replacement list 520 is created for all unique variables (step 501).

(3) Next, it is defined in the target server information definition 231 that is held in the system in the processing of the target server information definition at step 202 of the flow described with reference to FIG. It is confirmed whether or not the server name to be executed is included in the server name in the replacement list 520 held in the system by the processing in step 501 described above (step 502).

(4) If the presence of a server name that is not included in the replacement list 520 is confirmed in the confirmation in step 502, the command cannot be replaced in a server that is not included in the replacement list 520. It cannot be executed, and the process here is terminated as an abnormal termination or abnormal termination (step 503).

(5) When the confirmation in step 502 did not confirm the existence of a server name that is not included in the replacement list 520, or in the confirmation in step 500, it was confirmed that the command does not include an independent variable. In this case, the variable substitution process, which is the process here, is terminated, and the process is returned to the flow shown in FIG. 2 (step 51).

  Here, returning to the reference to the flow shown in FIG. 2, the processing after the command transmission processing in step 205 will be described. First, the command transmission process in step 205 will be described.

  In step 205, the command transmission process determines the server to which the command is to be sent with reference to the target server information definition 237 in which the server to be executed is defined, and the command to be executed is defined. The command to be transmitted is determined by referring to the execution command information definition 236 that is to be replaced, the server name to be replaced with the unique variable, and the replacement list 235 in which the replacement character string of the server name is held, and the determined server is determined. It is a process to send the command.

  The server in the server group 21 that receives the transmitted command and executes the command is executed by the operating system 22 operating in the target server in the command execution process in step 206. Each command transmitted to each target server in the transmission process is executed. The execution of the command depends on the function of the operating system, and is excluded from the scope of the present invention.

  Next, an execution result reception process for receiving the command execution result by the server in step 206 is performed as the process in step 207. The execution result transmission source server received in this execution result reception process is compared with the server name of the server list 230 and the verification flag, and the verification flag is set in the server name of the transmission source server included in the server list 230. It is confirmed by a confirmation process for confirming whether or not it is executed in the verification server in step 208.

  If it is confirmed in step 208 that the verification flag has been set, it is determined that the command execution in step 206 is the verification execution by the verification server, and the processing is shifted to the verification result confirmation 210. If the verification flag confirms that the verification flag is not set, it is determined that the command execution in step 206 is not the verification execution by the verification server but the execution other than the verification server, and the process is shifted to the difference processing 209. .

  FIG. 6 is a flowchart for explaining the details of the processing operation in the verification result confirmation unit 127, which will be described next. The processing operation in the verification result confirmation unit 127 described here is processing executed when it is determined in the verification server execution determination processing in step 208 in the flow shown in FIG. 2 that the execution is in the verification server. This is verification result confirmation processing in step 210 for determining whether the verification execution result in the verification server is normal or abnormal.

(1) When it is determined in step 208 of the flow described with reference to FIG. 2 that the execution is performed by the verification server, the process is transitioned to start the verification result confirmation process, First, an end code confirmation command is transmitted to the verification server that is the transmission source of the verification execution result received in step 207 of the flow illustrated in FIG. 2 in order to confirm the verification execution end code. In general, in a UNIX (registered trademark) operating system, if the end code is 0, the termination is normal, and if it is not 0, the termination is abnormal. Therefore, this characteristic is also used in the present invention to distinguish between normal and abnormal. However, in the case of a command that does not conform to this characteristic, it is impossible to make a determination according to the present invention, so that correction on the command side is required (steps 60 and 600).

(2) A response to the end code confirmation command transmitted to the verification server is received as a result of the end code confirmation command, and it is determined whether the received end code is 0, that is, normal, or other than 0, that is, abnormal ( Steps 601, 602).

(3) If it is determined in step 602 that the end code is abnormal, it is determined that the verification execution command in the present invention is abnormal, and the abnormal command is not executed except by the verification server. Abnormal termination processing is performed, and the processing here is terminated (step 603).

(4) If it is determined in step 602 that the end code is normal, it is determined that the verification execution command in the present invention is normal and the normal command is executed by a server other than the verification server. The process of verifying the verification result is completed, and the process is returned to the flow described with reference to FIG. 2 (step 61).

  Note that the processing returned to FIG. 2 transitions to the second connection 306 in the processing flow in the target server information definition unit shown in FIG. 3 for explaining the details of the target server information definition 202. By the subsequent step of the second connection 306, a server other than the verification server is held in the system as the target server in the target server information definition 321. As a result, the present invention allows the command to be executed by a server other than the verification server only when the command is determined to be normally executed.

  FIG. 7 is a flowchart for explaining the details of the processing operation in the difference processing unit 128, which will be described next. The processing operation in the difference processing unit 128 described here is processing that is executed when it is determined in the verification server execution determination processing in step 208 in the flow shown in FIG. 2 that it is not execution in the verification server. This is a difference process in step 209 for comparing the difference between the results of executing a command determined to be a normal command by the verification result confirmation of the work other than the verification server and determining whether the command is normal or abnormal.

(1) When it is determined in step 208 of the flow illustrated in FIG. 2 that the execution is performed by a server other than the verification server, the process is shifted to start the abnormal process. It is determined whether the execution target server that is the transmission source of the verification execution result received in step 207 of the flow shown in FIG. 2 is a single server or a plurality of servers (step 700).

(2) If it is determined in step 700 that there are a plurality of execution target servers, the difference in the execution result of each server can be compared, so that it is received in step 207 of the flow illustrated in FIG. The execution results of the servers to be executed are compared with each other, and the difference between the execution results is obtained (step 701).

(3) Next, as a result of the comparison in step 701, it is determined whether or not a difference is detected in each execution result, and if a difference is detected, whether the difference is abnormal or normal is detailed. Therefore, it is determined whether or not the detected difference is influenced by server-specific information (server name, IP address, date, time, user name) (step) 702, 703).

(4) If it is determined in step 703 that the difference is not influenced by server-specific information, it is necessary to regard the difference as abnormal, and there is a difference in the business command as the system of the present invention. And the difference is recognized as abnormal (step 708).

(5) Next, in order to make it easy for the worker to visually grasp the abnormal part, the part of the difference is displayed in different colors. At this time, if there are a plurality of different portions, different color coding is performed at each portion so that the worker can easily distinguish the plurality of different portions (step 709).

(6) After that, since it is determined that the difference detected in the determination in step 702 is abnormal, there is a possibility that an error will be induced by a subsequent command if the business is continued, so the system of the present invention is abnormally terminated. (Step 710).

(7) If it is determined in step 703 that the difference is affected by server-specific information, the difference is not regarded as abnormal, and there is a difference in the business command. Since it is influenced by the specific information, the difference is not recognized as abnormal (step 704).

(8) Next, in order to make it easier for the worker to visually grasp the difference location affected by the server-specific information, the location of the difference is displayed in different colors. At this time, if there are a plurality of different portions, different colors are used at the respective portions so that the worker can easily distinguish the plurality of different portions (step 705).

(9) After the processing of step 705, when the execution target server is a single unit in the determination in step 700 and the difference in the execution results of the servers cannot be compared, or in the determination in step 702 If no difference is detected between the results and it can be determined that the command has been executed normally, the command list 232 held in the execution command information definition process in step 203 in the flow illustrated in FIG. It is checked whether or not there is a subsequent command (step 706).

(10) If it is determined in step 706 that there is no subsequent command in the command list 232, the processing here is normally terminated. Thereby, a series of processes in the present invention is completed (step 711).

(11) If it is confirmed in step 706 that a subsequent command exists in the command list 232, the command pointer indicating the command position being executed in the command list 232 is advanced by 1, and the processing is shifted to the next command. . As a result, the process can be shifted to a process for executing the next command only when the command is normally completed (step 707).

(12) Then, the difference processing here is terminated, and the processing is returned to the flow shown in FIG. 2 is returned to the first connection 305 in the flow shown in FIG. 3 for explaining the details of the processing of the target server information definition unit in step 202, that is, the target server information definition processing in step 202. Transition. The target server information definition 321 is reset by the subsequent step of the first connection 305. As a result, when the verification flag exists, the verification server is held in the target server information definition 321 again (step 71). ).

  Each processing in the embodiment of the present invention described above can be configured as a processing program, and this processing program is stored in a recording medium such as HD, DAT, FD, MO, DVD-ROM, and CD-ROM. Can be provided.

  According to the above-described embodiment of the present invention, batch execution is possible when individual servers constituting a server group perform the same work, and even for a large-scale system, the work manpower for one server can be reduced. Work man-hours can be reduced, human error intervening points can be reduced, and key input for each server, which has conventionally been required, can be eliminated.

  Also, according to the embodiment of the present invention, differences in server environment can be absorbed by unique variables, so that work commands can be flexibly registered and automation can be easily performed.

  In addition, according to the embodiment of the present invention, the difference between work performed by a plurality of servers is automatically determined and color-coded, so that the workload is reduced, the work is accurate, and an abnormality occurs. It is possible to speed up the response, and it is possible to eliminate the need for the worker to visually check the execution result of each server and to find an abnormality.

  For example, in a system composed of 100 servers, when a certain package is introduced, commands can be collectively executed on 100 devices, and the number of man-hours can be simply reduced to 1/100. Further, by determining the difference in execution results, it becomes possible to automatically detect an abnormality, and further man-hour reduction can be expected.

  Since the embodiment of the present invention can systematically discriminate the difference between the execution results of a plurality of servers, it is much more accurate and quicker than the unreliable discrimination method that relies on human vision. The result can be derived.

DESCRIPTION OF SYMBOLS 10 Client machine 11 Network 12 Server group 100 Abnormality detection system 110 Registration part 111 Database registration process part 112 Business registration database 113 Process registration database 114 Server group registration database 115 Server registration database 116 Variable substitution registration database 120 Execution part 122 Target server information definition Section 123 Execution command information definition section 124 Variable substitution processing section 127 Verification result confirmation section 128 Difference processing section

Claims (2)

An error in the command execution result that executes a business process by executing a command on each server of a server group composed of multiple servers, determines the difference between the execution results of each server, and detects an abnormality in the business process In the detection system,
As database, server group registration database, process registration database, variable substitution registration database,
As a verification server and a production server that are servers to be executed for server business processing in which a server name and a verification flag are read from the server group registration database and the verification flag is set for the servers that constitute the server group The target server information definition section to be defined;
An execution command information definition unit that reads a command corresponding to a business process name from the process registration database and defines a command to be executed by the business process;
Variable replacement processing that reads a server name and a replacement character string from the variable replacement registration database, and replaces the unique variable included in the execution target command defined by the execution command information definition unit with a replacement character string corresponding to each server And
A verification result confirmation unit that determines normal termination and abnormal termination in the execution result of the command verified by the verification server;
An abnormality detection, comprising: a difference processing unit that compares execution results of each server in the execution results of commands executed by the production server to detect a difference and discriminates between normal termination and abnormal termination system. An error in the command execution result that executes a business process by executing a command on each server of a server group composed of multiple servers, determines the difference between the execution results of each server, and detects an abnormality in the business process In the detection method,
The database includes a server group registration database, a process registration database, and a variable substitution registration database.
As a verification server and a production server that are servers to be executed for server business processing in which a server name and a verification flag are read from the server group registration database and the verification flag is set for the servers that constitute the server group Define
Read the command corresponding to the business process name from the process registration database and define the command to be executed as the business process.
Wherein the variable substituents registration database reads and replacement strings and server names, performs replacement processing unique variables included in the execution target command defined in the replacement string corresponding to each server by executing the command information definition unit,
Determine the normal end and abnormal end in the execution result of the command verified by the verification server,
An abnormality detection method comprising: comparing the execution result of each server in the execution result of a command actually executed by the production server, detecting a difference, and discriminating between normal end and abnormal end.