JP5446649B2 - TAT measuring device, TAT measuring method and TAT measuring program - Google Patents

Description

  The present invention relates to monitoring a response of a web service, and more particularly to measuring a TAT (Turnaround Time) of the service.

  The web service provided by the web server needs to continue to operate 24 hours a day and without stressing the user. For this reason, not only the occurrence of a failure but also a decrease in performance due to, for example, the concentration of access to a specific process must be discovered immediately.

  For this reason, for example, pseudo access (hereinafter simply referred to as “access” in this specification) that assumes screen transition of an end user's web service, such as NEC's WebSAM (registered trademark) application navigator, is used as a web service. There are products that do this and thus measure TAT.

  In this TAT measurement, for example, in “shopping at an online shop”, a series of screen transitions such as login → product selection → confirmation → confirmation is defined as a “scenario”, and a user connects to a web server via a web browser The operation that will be performed is performed as an access process, and the TAT for it is measured.

  There are the following technical documents related to the measurement of TAT for web services. Patent Document 1 describes a measurement system that can measure response time even for a web service that requires a password input. Patent Document 2 describes scheduling for assigning jobs to a plurality of computers.

  Patent Document 3 describes that a “scenario” executed by a web service is divided and executed. Patent Document 4 describes a response time measurement monitor that measures a response time that a user actually experiences in a web service. Non-Patent Document 1 describes a specific example of a product that measures the TAT of a web service (refer to the item “Web scenario operation monitoring” in the “Probe option” column).

JP200002915475A JP 2005-031771 A JP 2008-250953 A Japanese Patent Laid-Open No. 11-110312

NEC Corporation, “WebSAM Application Navigator-Features / Functions”, [online] [searched July 24, 2009], Internet <URL: http://www.nec.co.jp/middle/WebSAM/products /appnavi/kinou.html>

  In such TAT measurement, when one measuring device executes a plurality of scenarios, a timing at which a plurality of access processes are simultaneously executed may occur. When a plurality of processes are executed simultaneously, processing overhead occurs in the apparatus itself, response response is delayed, and accurate TAT measurement may be difficult.

  It may be possible to prevent a plurality of access processes from being simultaneously executed by one measuring device by optimizing the schedule. However, since the scenario to be measured may be several hundred to several thousand units depending on the case, it is difficult to optimize the schedule in such a case.

  Furthermore, it is conceivable to prepare a plurality of measuring devices and allocate a plurality of scenarios to each device so that they can be executed simultaneously. However, the operation shown in the above scenario is originally performed by one terminal operated by the user. Therefore, especially for web services that need to store IDs, passwords, etc. in cookies, it is impossible to divide a scenario into steps and have each measurement device execute them. Even if a measuring device is prepared, the load on access processing cannot be distributed effectively.

  Patent Documents 1 to 4 and Non-Patent Document 1 describe a technique for measuring TAT from a single scenario, particularly a technique for measuring TAT for a web service that requires input of a password or the like. There is no description of a configuration that can solve the problem that occurs when such a plurality of scenarios are executed simultaneously. In particular, there is no description of a configuration that makes it possible to measure a TAT by dividing a plurality of scenarios for each step.

  An object of the present invention is to provide a TAT measurement apparatus, a TAT measurement method, and a TAT measurement program that can accurately measure a TAT by effectively distributing a load applied to an access process.

  In order to achieve the above object, a TAT measuring apparatus according to the present invention is a TAT measuring apparatus that measures a TAT (Turnaround Time) of a web server in response to an operation by a user, and generates a plurality of operations by a plurality of users in a pseudo manner. A response that is cookie information corresponding to the operation content to be executed by the request generation unit and each stage of the operation content, and a control unit that controls operations of the plurality of request generation units. Response information table that records parameters, empty search part that searches for empty ones among multiple request generation means, and operation contents and response parameters for each stage are assigned to empty request generation means Mediating the access to the execution instruction unit to be executed and the web server generated by the request generation means. This measures the time from Seth until the response to this is returned, characterized in that a proxy server unit to TAT.

  In order to achieve the above object, the TAT measurement method according to the present invention includes a plurality of request generation means for generating pseudo operations by a plurality of users, and corresponds to each operation content to be executed and each operation content stage. A TAT (Turnaround Time) measuring device that stores in advance a response parameter that is cookie information to be measured, the TAT measuring method for measuring the TAT of a web server in response to an operation by a user, and a plurality of request generating means Search for an empty request, assign the operation contents and response parameters for each stage to the empty request generation means, execute, and mediate the access to the web server generated by the request generation means. The time until the response to the response is returned is measured and this is set as TAT.

  In order to achieve the above object, a TAT measurement program according to the present invention includes a plurality of request generation means for generating pseudo operations by a plurality of users, and corresponds to each operation content to be executed and each operation content stage. A TAT (Turnaround Time) measuring device that stores in advance response parameters that are cookie information to be searched, and searches for a free one of a plurality of request generating means in a computer provided in the TAT measuring device. A procedure, a procedure for assigning an operation content and a response parameter for each stage to an empty request generation means, a procedure for mediating access to the web server generated by the request generation means, and a response to the response from the access Measuring the time until a reply is made and executing this procedure as TAT To.

  As described above, since the present invention is configured to include the response information DB storing the response parameters necessary for each step, it is possible to allocate and execute processing for each step to a plurality of request generation means. Accordingly, it is possible to provide a TAT measuring apparatus, a TAT measuring method, and a TAT measuring program having an excellent feature that it is possible to measure an accurate TAT by dividing a large number of scenarios for each step.

It is explanatory drawing which shows the structure of the whole response monitoring apparatus which concerns on embodiment of this invention. It is explanatory drawing which shows the more detailed structure as a software of the request production | generation means and control means shown in FIG. It is explanatory drawing shown about the memory content of the scenario attribute table and request production | generation means attribute table contained in the monitoring DB shown in FIG. It is explanatory drawing shown about the memory | storage content of the response information table and measurement result table which are contained in monitoring DB shown in FIG. It is explanatory drawing which shows an example of the content of the scenario defined with the response monitoring apparatus shown in FIGS. It is a flowchart which shows operation | movement when the control means shown in FIGS. 1-2 is the operation | movement which allocates the scenario which shows operation | movement content to a request production | generation means, and the step is an initial step. FIG. 7 is a flowchart illustrating an operation performed by the request generation unit and the TAT monitoring unit that have received the contents of a step waiting for execution in step S <b> 408 in FIG. 6, when the step is an initial step. FIG. 9 is a flowchart showing operations executed by the request generation unit and the TAT monitoring unit when the contents of an intermediate step waiting for execution are transferred from the scenario execution instruction unit 240 following the operation of FIG. 7. FIG. 9 is a flowchart showing empty request search and intermediate step execution instruction operations shown as step S602 in FIG. 8; FIG. FIG. 9 is a flowchart showing operations executed by the request generation unit and the TAT monitoring unit when the contents of the final step waiting for execution are delivered from the scenario execution instruction unit following the operations of FIGS. It is explanatory drawing which shows an example of operation | movement in which the control means actually allocates a several scenario to each of several request production | generation means in step unit with the response monitoring apparatus demonstrated so far.

(First embodiment)
Hereinafter, the configuration of an embodiment of the present invention will be described with reference to FIGS.
First, the basic content of the present embodiment will be described, and then more specific content will be described.
The TAT measurement device (response monitoring device 10) according to the present embodiment is a device that measures a TAT (Turnaround Time) of a web server in response to an operation by a user. This apparatus includes a plurality of request generation units 110 to 130 that artificially generate operations by a plurality of users, and a control unit 200 that controls operations of the plurality of request generation units. The control means includes a response information table 293 that records the operation contents to be executed by the request generation means and response parameters that are cookie information corresponding to each stage of the operation contents, and a plurality of request generation means. An empty search unit 230 that searches for an empty one, an execution instruction unit (scenario execution instruction unit 240) that assigns an operation content and a response parameter for each stage to an empty request generation unit, and executes the request generation unit. A proxy server unit (TAT monitoring unit 260) that mediates the access to the generated web server, measures the time from the access until a response to the response is returned, and sets this as TAT.

  This proxy server unit records the occurrence time of the access from the request generation means to the web server and the occurrence time of the response from the web server to the request generation means, and obtains the TAT from these times. H.263. Further, the execution instruction unit causes the request generation unit to check the response from the web server for the operation result of the step immediately before the step immediately before causing the request generation unit to execute the operation contents for each step. It has a function.

  The control unit further temporarily stores the operation contents for each stage in the prepared scenario execution queue 251 when there is no available request generation unit among the plurality of request generation units. A scenario standby unit 250 is provided for causing the vacant request generation means to execute the operation content for each stage as soon as a vacancy occurs therein. The control unit also includes a scenario recorder unit 210 that records operations performed by the user on the web server and stores them as operation contents for each stage.

  The control unit also includes a request generation unit attribute table 292 that stores the IP address of each of the plurality of request generation units and the usage status indicating whether each of the plurality of request generation units is in use. . The execution instruction unit has a function of notifying the proxy server unit of the request generation unit to which the operation content for each step is allocated when the operation content for each step is allocated to the request generation unit.

  The request generation unit 110 includes a scenario execution instruction receiving unit 111 that receives an execution instruction for the operation content for each step from the execution instruction unit, and a request transmission unit 112 that transmits a request based on the operation content for each step to the web server. A response reception unit 113 that receives a response from the web server to the request, a response screen display unit 114 that displays the response on the screen, and determines whether or not the response is a valid response and displays the determination result in the response information table And a response information transmission unit 115 to be recorded.

By providing this configuration, the TAT measuring apparatus can measure an accurate TAT by dividing a large number of scenarios for each step.
Hereinafter, this will be described in more detail.

  FIG. 1 is an explanatory diagram showing the overall configuration of the response monitoring apparatus 10 according to the embodiment of the present invention. The response monitoring device 10 is configured by a normal computer device, and is connected via the Internet 20 to web servers 30 and 31 that provide web services to be monitored.

  However, the response monitoring apparatus 10 may be a single computer apparatus or a plurality of computer apparatuses. In the present embodiment, an example in which the response monitoring apparatus 10 is a single computer apparatus is shown. Further, although only two web servers 30 and 31 to be monitored are shown in FIG. 1, any number of web servers 30 and 31 may be provided. Hereinafter, these are collectively referred to as the web server 30 unless otherwise required.

  The response monitoring apparatus 10 includes a processor 11 that executes a computer program, a storage unit 12 that stores a computer program and data, an input / output unit 13 that inputs and outputs data, and another computer via the Internet 20. And communication means 14 for performing data communication with the apparatus.

  In the processor 11, the first to third request generation means 110, 120, and 130 and the control means 200 are executed as computer programs. In the present embodiment, an example is shown in which three request generation units are executed, but in actuality, any number of two or more request generation units may be operating. The control unit 200 controls and operates the first to third request generation units 110 to 130. The storage unit 12 stores a monitoring DB (database) 290 including each table described later and a scenario execution queue 251 described later.

  FIG. 2 is an explanatory diagram showing a more detailed configuration of the first to third request generation means 110 to 130 and the control means 200 shown in FIG. 1 as software. The arrows in FIG. 2 conceptually show the flow of information between the parts.

  The request generation unit 110 includes functional units such as a scenario execution instruction reception unit 111, a request transmission unit 112, a response reception unit 113, a response screen display unit 114, and a response information transmission unit 115. Similarly, the request generation means 120 and 130 include the same functional units as the request generation means 110, and operate under the same control from the control means 200. Since the configuration of the request generation means 120 and 130 is the same as the configuration of the request generation means 110, detailed description is omitted in FIG.

  The request generating means 120 includes a scenario execution instruction receiving unit 121... And the request generating unit 130 includes a scenario execution instruction receiving unit 131. The order is “12”, and the units included in the request generation unit 130 are the hundreds and tens of the reference number “13”.

  The scenario execution instruction receiving unit 111 (121, 131) receives an instruction regarding the operation content from a scenario execution instruction unit 240 of the control unit 200 described later. The request transmission unit 112 (122, 132) causes the web server 30 to generate pseudo access according to the instructed operation content. The response receiving unit 113 (123, 133) receives a response from the web server 30 for this pseudo access.

  The response screen display unit 114 (124, 134) is a web browser, and displays the response received from the web server 30 as the screen of the web browser. The response information transmitting unit 115 (125, 135) confirms whether or not the response of the screen displayed on the web browser is a correct screen in which no error has occurred, and the control unit 200 described later with the confirmation result. Are sent to the DB control unit 220 for recording.

  The control unit 200 includes functional units such as a scenario recorder unit 210, a DB control unit 220, an empty search unit 230, a scenario execution instruction unit 240, a scenario standby unit 250, a TAT monitoring unit 260, a response information monitoring unit 270, and an execution timer unit 280. Is provided.

  The scenario recorder unit 210 records the operation content for the website input from the user via the input / output means 13 and stores it in the DB control unit 220 as a scenario to be described later. The DB control unit 220 refers to and updates the storage contents of each table such as the scenario attribute table 291, the request generation unit attribute table 292, the response information table 293, and the measurement result table 294 included in the monitoring DB 290 stored in the storage unit 12. Do.

  The empty search unit 230 searches for empty ones among the first to third request generation units 110 to 130. The scenario execution instructing unit 240 causes the empty request generating unit searched by the empty searching unit 230 to execute the operation of the step to be executed in the scenario. The scenario standby unit 250 stores the operation of the step waiting for execution in the first to third request generation units 110 to 130 in the scenario execution queue 251 in the storage unit 12 by a queue (first-in first-out) method. The first to third request generation units 110 to 130 are made to execute as soon as they are available.

  The TAT monitoring unit 260 has a function as a proxy server, mediates pseudo access to the web server 30 from the first to third request generation units 110 to 130, and acquires an occurrence time of this access. A response mediation function unit 261 that mediates a response from the web server 30 to the first to third request generation units 110 to 130 with respect to the pseudo access, and acquires an occurrence time of the response, an access mediation function unit 262, And a TAT calculation function unit 263 that calculates TAT from the time acquired by the response mediation function unit 261. The calculated TAT is recorded in the response information table 293 via the DB control unit 220.

  The response information monitoring unit 270 analyzes the recorded TAT, and outputs this TAT to the input / output unit 13 together with the analysis result. The analysis here refers to, for example, detecting a case where the TAT is equal to or greater than a specific threshold or an error has occurred on the response screen, and issuing an alarm to the administrator. The execution timer unit 280 counts an execution interval such as once per minute, for example, and executes processing described later.

  3 to 4 are explanatory diagrams showing storage contents of the scenario attribute table 291, the request generation unit attribute table 292, the response information table 293, and the measurement result table 294 included in the monitoring DB 290 shown in FIG. 2. The scenario attribute table 291 shown in FIG. 3 stores contents such as a scenario name 291a, an execution interval 291b, a next execution start time 291c, and a step name 291d. The request generation unit attribute table 292 shown in FIG. 3 also stores the contents such as the request generation unit name 292a, the IP address 292b of the request generation unit, and the current usage status 292c.

  Here, the execution interval 291b is recorded content that the scenario recorder unit 210 stores via the DB control unit 220 based on user input, as will be described later. On the other hand, the next execution start time 291c is the recorded content that is subtracted by the DB control unit 220 with the passage of time. When this becomes “0”, the execution of the scenario corresponding thereto is started.

  The response information table 293 illustrated in FIG. 4 stores contents such as a scenario name 291a, a step name 291d, a pre-transition URL (Uniform Resource Locator) 293a, a response parameter 293b, and a post-transition URL 293c. The scenario name 291a and the step name 291d correspond to the data of the same name in the scenario attribute table 291. The pre-transition URL 293a and the response parameter 293b are data necessary when the first to third request generation units 110 to 130 connect to the web server 30 via the control unit 200 and perform processing. When the processing is completed, the screen moves to the post-transition URL 293c screen. In other words, if the screen moves to the post-transition URL 293c screen after the processing is completed, the processing is successful.

  In the measurement result table 294 also shown in FIG. 4, the contents such as the scenario name 291a, the step name 291d, and the measurement result 294a are stored. The scenario name 291a and the step name 291d correspond to the data of the same name in the scenario attribute table 291. In addition, the measurement result 294a includes one or a plurality of sets of the date / time 294b when the measurement is performed and the TAT 294c measured at the date / time.

  FIG. 5 is an explanatory diagram showing an example of the contents of the scenario defined by the response monitoring apparatus 10 shown in FIGS. For example, assuming a scene where the user performs “shopping at a certain net shop”, the user who first accesses the web server 30 and opens the login screen 301 inputs the ID and password here and clicks the “Send” button. Push. Then, the display screen proceeds to the product selection screen 302.

  Subsequently, the user selects a product number on the product selection screen 302 and presses a “select this product” button. Then, the display screen proceeds to the confirmation screen 303. Subsequently, the user inputs the purchase points on the confirmation screen 303 and presses the “confirm” button. Then, the display screen proceeds to the confirmation screen 304. Finally, the user inputs a delivery address on the confirmation screen 304 and presses the “confirm” button. Thus, the purchase of the product is confirmed, a series of operations are completed, and a completion screen 305 is displayed. The contents input by the user in the process so far are stored in the cookie of the web browser.

  The user performs this operation on the terminal of the response monitoring apparatus 10 in advance, and the scenario recorder unit 210 records the operation performed by the user. More specifically, input contents such as IDs, passwords, product numbers, purchase points, delivery addresses, etc. entered by the user, input element names such as buttons operated by the user such as “Send”, “Confirm”, etc. The URL and the contents stored in the cookie of the web browser in the course of operation.

  The scenario recorder unit 210 stores the scenario attribute table 291 and the response information table 293 via the DB control unit 220. The scenario name 291a can be arbitrarily defined by the user, for example, “shopping at the net shop XX”. The step name 291d is a sequential number given for each operation step, for example.

  In the example shown in FIGS. 3 to 5, the scenario name is “A”, the transition from the login screen 301 to the product selection screen 302 is the step name “A1”, and the transition from the product selection screen 302 to the confirmation screen 303 is performed. A step name “A2”, a transition from the confirmation screen 303 to the confirmation screen 304 is defined as a step name “A3”, and a transition from the confirmation screen 304 to the completion screen 305 is defined as a step name “A4”. The first step executed in each scenario is referred to as a first step, the last step executed is referred to as a final step, and a step that does not correspond to the first step or the final step is referred to as an intermediate step. The first step of the scenario name “A” is “A1”, the final step “A4” is the intermediate steps “A2” and “A3”.

  The URL of each screen is stored as the response parameter 293b of the response information table 293, the URL 293a before the transition of the response information table 293, the input ID and password, the contents of the cookie, and the like. Then, when the user inputs the above scenario and the time to perform these processes defined as each step, the scenario recorder unit 210 stores the execution interval 291b of the scenario attribute table 291 via the DB control unit 220. .

  FIG. 6 shows an operation in which the control unit 200 shown in FIGS. 1 and 2 allocates a scenario indicating the operation content to the first to third request generation units 110 to 130, and that step is an initial step. It is a flowchart. In the scenario attribute table 291 and the response information table 293, the scenario recorder and the request generation unit attribute described above are stored in advance by the scenario recorder unit 210.

  First, the execution timer unit 280 counts the execution interval (step S401), and when the execution interval comes, the DB control unit 220 updates the next execution start time 291c (step S402). In step S402, the next execution start time 291c is subtracted with the passage of time.

  Then, the DB control unit 220 determines whether there is a scenario in which the next execution start time 291c is 0 for each scenario stored in the scenario attribute table 291 (step S403). First, it is assumed that “step A1”, which is the first step of “scenario A” described above, is an execution target.

  If there is a scenario in which the next execution start time 291c becomes 0, this is a scenario that should be started immediately, so that the empty search unit 230 uses the request generation means attribute table 292 usage status 292c for the first step. Referring to FIG. 5, it is determined whether or not there are currently unused (empty) request generation units 110 to 130 (step S404). Here, it is assumed that, among the first to third request generation units 110 to 130, the first request generation unit 110 is free.

  If there is a vacancy, the DB control unit 220 sets the current usage status 292c of the first request generation unit 110 in the request generation unit attribute table 292 to “in use” (step S405). At the same time, the DB control unit 220 reads the content of step A1 and passes it to the scenario execution instruction unit 240 (step S408). Then, the DB control unit 220 updates the next execution start time 291c for the scenario (step S409).

  In step S404, if there is no free space in the first to third request generation units 110 to 130, the scenario standby unit 250 stores the contents of step A1 to be executed in the scenario execution queue 251 (step S406). The contents of step A1 stored here are executed as soon as any one of the first to third request generation units 110 to 130 becomes available (step S407) and proceeds to step S405.

  In step S408, the content of step A1 waiting for execution is sent to the scenario execution instruction unit 240, and at the same time, information that the request generation unit 110 executes step A1 is transmitted from the scenario execution instruction unit 240 to the TAT monitoring unit 260. (Step S410).

  The job that has been processed in steps S406 to S407 and stored in the scenario execution queue 251 is executed after waiting for any of the first to third request generation units 110 to 130 to be executed. The job is executed with a slight delay from the time set as the interval 291b. For example, if the set time of the execution interval 291b has elapsed while “waiting for a free space”, the same scenario is repeatedly executed. In order to avoid such a situation, in step S409, the next execution start time 291c subtracted with the passage of time is returned to the value of the execution interval 291b. In other words, the next execution start time 291c is “reset” to the set value of the execution interval 291b.

  FIG. 7 is a flowchart showing the operation performed by the request generation unit 110 and the TAT monitoring unit 260 that have received the contents of the step waiting for execution in step S408 of FIG. 6 and the step is an initial step. Here, it is assumed that the content of step A1 waiting for execution is transferred from the scenario execution instruction unit 240 to the first request generation unit 110 and executed.

  The scenario execution instruction receiving unit 111 that has received the content of step A1 first activates the response screen display unit 114 that is a web browser (step S501), and displays the pre-transition URL 293a (here, the login screen 301) as in the scenario. open.

  Then, the scenario execution instruction receiving unit 111 passes the pre-transition URL 293a and the response parameter 293b to the request transmitting unit 112, and transmits a login request (step S502). This request is transmitted to the web server 30 via the TAT monitoring unit 260 which is a proxy server.

  When the access mediation function unit 262 of the TAT monitoring unit 260 receives this request (step S503), the TAT calculation function unit 263 starts counting TAT (step S504), and the access mediation function unit 262 receives this request from the web server 30. (Step S505). Since the processing of the request generation unit 110 ends once so far, the TAT monitoring unit 260 sets the usage status 292c of the request generation unit 110 to “unused” via the DB control unit 220 (step S506).

  The web server 30 starts executing a process (transaction) corresponding to the request transferred by the access mediating function unit 262 (step S507).

  FIG. 8 shows the first to third request generation means 110 to 130 when the contents of the intermediate step waiting for execution (here, step A2) are transferred from the scenario execution instruction unit 240 following the operation of FIG. 4 is a flowchart illustrating an operation performed by the TAT monitoring unit 260. As described above, the request of step A1 has already been transmitted by the first request generation means 110. The web server 30 completes the process for step A1, and tries to return a response to this to the request generation unit 110 (step S601).

  This result is returned to one of the request generation units 110 to 130 via the response mediation function unit 261 of the TAT monitoring unit 260 that is a proxy server. Here, it is assumed that the request generation unit 120 different from the request generation unit 110 that has transmitted the request of Step A1 is free.

  When the response mediation function unit 261 of the TAT monitoring unit 260 receives this request, the TAT calculation function unit 263 stops counting the TAT (step S602), and the empty search unit 230 includes the empty request generation units 110 to 130. It is determined whether or not (step S603). FIG. 9 is a flowchart showing the empty search of the first to third request generation means 110 to 130 shown as step S602 in FIG. 8 and the operation of executing an intermediate step (here, step A2). Since the content of this operation is the same as the operation of steps S404 to S410 in FIG. 6, the reference number of each step is indicated by +50. However, there is no operation of updating the next execution start time 291c for the scenario corresponding to step S409 in FIG.

  Here, following the operation in step S603 (FIG. 8), the response mediation function unit 261 transfers this request to the request generation unit 120 (step S604). The count of the time started in step S504 in FIG. 7 and stopped in step S602 in FIG. 8 is the TAT for step A1.

  In step S410 of FIG. 6, since the information that the request generation unit 110 transmits the request for step A1 has already been transmitted, the TAT calculation function unit 263 has already grasped this. Further, the TAT calculation function unit 263 refers to the IP address 292b of the request generation unit attribute table 292, and grasps that the request corresponding to the IP address 292b is from the request generation unit 110. Furthermore, it is understood that the response should be returned to the request generation unit 120 by the operation in step S460 of FIG.

  When the TAT calculation function unit 263 can count the TAT related to step A1, this time is transmitted to the DB control unit 220 (step S605). The response receiving unit 123 receives the response transferred from the response mediation function unit 261 in step S604 and displays it on the response screen display unit 124 (step S606). Then, the response information transmission unit 125 checks whether the response displayed on the screen is a correct response that is not an error screen or the like, and transmits the result to the DB control unit 220 (step S608). Here, whether the displayed URL matches the post-transition URL 293c in step A1 can be used as a determination material, but the determination material is not limited to this.

  Following this, the request generation unit 120 performs an operation related to the transmission of the request relating to Step A2 that has received the execution instruction from the control unit 200 (Steps S608 to 613). The content is the same as the content described in steps S502 to S507 in FIG. 7, and only the subject to be executed is changed from the request generation unit 110 to the request generation unit 120. The same operation is performed for the subsequent step A3.

  In this embodiment, the confirmation of the response screen of the previous step and the request transmission of the current step are performed in one set in this way. For example, if “now step” is step A2, “previous step” is step A1. In this way, for example, even if responses related to a plurality of scenarios transmitted by one request generation means are simultaneously returned, the response confirmation processing for those simultaneously returned responses is not allocated to the same request generation means. , It won't delay processing.

  FIG. 10 shows the first to third request generation means 110 to 110 when the contents of the final step waiting for execution (here, step A4) are passed from the scenario execution instructing unit 240 following the operations of FIGS. 13 is a flowchart illustrating operations executed by 130 and the TAT monitoring unit 260. Since the content of the operation is the same as that in steps S601 to 607 in FIG. 8 except that there is no operation corresponding to step S603 in which it is determined whether or not there are empty request generation units 110 to 130, these These steps are indicated as steps S701 to 707 in which the reference number of each step is +100. However, in step S708 after completion of each step, the response information transmission unit 115 (125, 135) sets the usage status 292c of the request generation unit 110 to “unused” via the DB control unit 220 (step S708). ).

  The DB control unit 220 records the transmitted TAT measurement result and response confirmation result as the measurement result 294 a of the measurement result table 294. The response information monitoring unit 270 constantly monitors and analyzes the measured TAT value and the presence / absence of an error. If there is an abnormality in the measured TAT, an alarm is immediately output to the input / output means 13. In other cases, the administrator can cause the input / output means 13 to output the measurement result and the determination result at any time by an operation. Any known method can be used for monitoring and analyzing the TAT. For example, the measured TAT may be compared with a predetermined threshold value, and an abnormality may be determined if the measured TAT is equal to or greater than the threshold value.

  FIG. 11 is an explanatory diagram illustrating an example of an operation in which the control unit 200 actually allocates a plurality of scenarios to each of the plurality of request generation units 110 and 120 in units of steps in the response monitoring apparatus 10 described so far. is there. The vertical direction in FIG. 11 shows the passage of time.

  Here, there are two scenarios, scenarios A and B. Scenario A is a series of steps A1, A2, A3, etc. described in FIG. 5, and scenario B is the same steps B1 and B2 as described above. A series of steps are executed in order. The scenario A is operated by connecting to the web server 30, and the scenario B is connected to another web server 31.

  In order to avoid the confusion of the drawings, in FIG. 8, the request generation means are the request generation means 110 and 120, and the scenarios and steps are only described in steps A1 to A3 and steps B1 and B2. A larger number of scenarios and steps can be allocated to a large number of request generation means.

  First, the control unit 200 allocates the execution of step A1 to the request generation unit 110 and transmits an execution instruction (step S800, step S408 in FIG. 6). Based on the allocated contents, the request generation unit 110 accesses the web server 30 to transmit a request, and performs an operation of proceeding from the login screen 301 to the product selection screen 302 (step S801, step S501 in FIG. 7). 502). At this time, the ID and password that need to be input use the contents stored in the response information table 293 as the response parameter 293b.

  Since the access to the web server 30 in step S801 passes through the control means 200 that is a proxy server, the control means 200 that has received this starts TAT counting here (step S504 in FIG. 7). Here, the control means 200 instructs the request generation means 110 whose use status 292c has become “unused” after transmitting the request of step A1 in step S801, this time to execute step B1 (step S802).

  Receiving this, the request generation means 110 accesses the web server 31 and transmits the request of step B1 (step S803). Here, between step S802 and step S803, a response to step A1 is returned from the web server 30, but the request generation means 110 is being used in the processing related to step B1.

  Therefore, the control unit 200 searches the request generation unit for free in the process of FIG. 9 and instructs the request generation unit 120 that has been free to execute the response to step A1 and the execution of the request to step A2. After stopping the TAT count, the response is transferred to the request generation means 120 (step S810, steps S603 to 604 in FIG. 8).

  Here, the TAT calculation function unit 263 acquires the time between step S801 and step S810 as the TAT related to step A1, and records it in the measurement result table 294. The request generation unit 120 to which the response is transferred from the control unit 200 returns the result of confirming whether or not this response screen is normal to the control unit 200 (step S811, steps S606 to 607 in FIG. 8).

  The control unit 200 confirms the response (that is, the product selection screen 302) of step A1 to the request generation unit 120 different from the request generation unit 110 that transmitted the request of step A1, and the subsequent step A2, that is, the product selection screen 302. The operation of transmitting a request to proceed to the confirmation screen 303 can be executed. This is because the response parameter 293 b necessary for proceeding from the product selection screen 302 to the confirmation screen 303 is stored in the response information table 293.

  Upon confirming the response in step A1, the request generation unit 120 receives an instruction from the control unit 200 to transmit the request in step A2, that is, to access the web server 30 and transmit the request. An operation of proceeding from 302 to the confirmation screen 303 is performed (steps S812 to 813). Since the access to the web server 30 in step S813 passes through the control unit 200 that is a proxy server, the control unit 200 that has received this starts the TAT count in step A2.

  Here, the response to the request in step B1 transmitted to the web server 31 in step S803 returns to the control unit 200 that is a proxy server (step S814). The response confirmation may be performed by either of the request generation means 110 and 120, but here, the request generation means 120 is assumed to perform the response, and the response is transferred to the request generation means 120 after the TAT count is stopped ( Step S815).

  The request generation unit 120 that has confirmed the response of Step B1 receives the instruction from the control unit 200 to transmit the request of Step B2, that is, to access the web server 31, and to transmit the request. This is performed (steps S816 to 817).

  Here, the response to the request in step A2 transmitted to the web server 30 in step S813 has returned to the control unit 200 that is a proxy server (step S804). The confirmation of this response may be performed by either of the request generation means 110 and 120, but here the request generation means 110 is assumed to stop the TAT count and transfer this response to the request generation means 110. The request generation unit 110 confirms this response and returns the result to the control unit 200 (step S805).

  The response monitoring apparatus 10 performs the above operation for each step following step A2, each step following step B1, and scenarios other than scenario A and scenario B. At that time, be sure to check the response screen of the previous step and send the request of the current step in one set. Similarly, the request generation means other than the request generation means 110 and 120 are assigned and executed for each step of each scenario.

  The TAT monitoring unit 260, which is a proxy server, uses an IP address (local address) that can be used only within an organization and an IP address (global address) on the Internet by using IP masquerade, which is a well-known technology defined in RFC1631 (May 1994). Address) transparently. For this reason, even if the request generation means different from the operation of step A1 executes the operation of step A2, and even if the request generation means different from sending the request and confirming the response thereto, This is not known, and the web server 30 recognizes only that a series of processing is continued for the same proxy server (TAT monitoring unit 260). Therefore, if there is no mistake in the response parameter 293b, the web server 30 can continue the processing without any problem.

  Since the IP address 292 of each of the first to third request generation units 110 to 130 is recorded in the request generation unit attribute table 292, the TAT monitoring unit 260 performs which request generation unit executes which step. As long as it can be ascertained (step S410 in FIG. 6), an accurate TAT can be measured. Of course, even if the request generation means that transmitted the request is different from the request generation means that confirmed the response to the request generation means, there is no particular effect on the TAT measurement.

(Overall operation of the first embodiment)
Next, the overall operation of the above embodiment will be described. The TAT measurement method according to the present embodiment includes a plurality of request generation units that artificially generate an operation by a plurality of users, and is an operation content to be executed and a response that is cookie information corresponding to each stage of the operation content This is a TAT (Turnaround Time) measuring device that stores parameters in advance, and is a TAT measuring method for measuring the TAT of a web server in response to an operation by a user. In this TAT measurement method, an empty one of a plurality of request generation means is searched (FIG. 6: Step S404), and an operation content and a response parameter are assigned to the empty request generation means and executed. (FIG. 6: Step S408), mediates the access to the web server generated by the request generation means (FIG. 7: Steps S503 to 505), and measures the time from the access to the response to the response. Is TAT (FIG. 8: step S603).

Here, each of the above-described operation steps may be programmed to be executable by a computer, and may be executed by an agent machine which is a computer that directly executes each of the steps.
With this configuration and operation, the present embodiment has the following effects.

  In the present embodiment, since operation of two or more steps is not simultaneously allocated to one request generation unit, processing overhead due to simultaneous execution of a plurality of access processes does not occur. Therefore, operations can be efficiently allocated to each request generation means in units of steps, and the usage rate can be increased.

  Since the operation of allocating the step operation to the request generation means is performed without human intervention, it is not necessary to manually schedule a complicated and difficult monitoring scenario execution order. Furthermore, it is possible to flexibly cope with changes in monitoring conditions and addition / deletion of services to be monitored without human intervention.

  The present invention has been described with reference to the specific embodiments shown in the drawings. However, the present invention is not limited to the embodiments shown in the drawings, and any known hitherto provided that the effects of the present invention are achieved. Even if it is a structure, it is employable.

  INDUSTRIAL APPLICABILITY The present invention can be used for web server operation management, in particular, prevention of trouble occurrence and suppression of performance degradation.

DESCRIPTION OF SYMBOLS 10 Response monitoring apparatus 11 Processor 12 Memory | storage means 13 Input / output means 14 Communication means 20 Internet 30, 31 Web server 110, 120, 130 Request generation means 111, 121, 131 Scenario execution instruction receiving part 112, 122, 132 Request transmission part 113 , 123, 133 Response reception unit 114, 124, 134 Response screen display unit 115, 125, 135 Response information transmission unit 200 Control unit 210 Scenario recorder unit 220 DB control unit 230 Free search unit 240 Scenario execution instruction unit 250 Scenario standby unit 251 Scenario execution queue 260 TAT monitoring unit 261 Response mediation function unit 262 Access mediation function unit 263 TAT calculation function unit 270 Response information monitoring unit 280 Execution timer unit 290 Monitoring B
291 Scenario attribute table 292 Request generation means attribute table 293 Response information table 294 Measurement result table

Claims (10)

A TAT measuring device for measuring a TAT (Turnaround Time) of a web server in response to an operation by a user,
A plurality of request generation means for artificially generating operations by the plurality of users, and a control means for controlling operations of the plurality of request generation means,
The control means is
A response information table that records the operation content to be executed by the request generation means and a response parameter that is cookie information corresponding to each stage of the operation content;
A free search unit for searching for a free one of the plurality of request generation means;
An execution instructing unit that causes the vacant request generating means to assign and execute the operation content for each stage and the response parameter;
A proxy server unit that mediates the access to the web server generated by the request generation unit, measures the time from the access until a response to the response is returned, and sets this as the TAT A characteristic TAT measuring device.   The proxy server unit records an occurrence time of access from the request generation means to the web server and an occurrence time of a response from the web server to the request generation means, and from these times, the TAT The TAT measuring apparatus according to claim 1, further comprising a TAT calculating function for acquiring the TAT.   Immediately before the execution instructing unit causes the request generation unit to execute the operation content for each step, the request generation unit confirms the response from the web server for the operation result of the step immediately before the step. The TAT measuring apparatus according to claim 1, further comprising a function to be executed by the TAT.   When the control means does not have any available request generation means, the operation contents for each stage are temporarily stored in a scenario execution queue prepared in advance, and the plurality of request generations are performed. 2. The TAT measurement apparatus according to claim 1, further comprising a scenario standby unit that causes the vacant request generation means to execute the operation content for each stage as soon as a vacancy occurs in the means.   The TAT measuring apparatus according to claim 1, wherein the control unit includes a scenario recorder unit that records an operation performed on the web server by a user and stores the operation content for each stage.   The control unit includes a request generation unit attribute table storing an IP address of each of the plurality of request generation units and a usage status indicating whether each of the plurality of request generation units is in use. The TAT measuring apparatus according to claim 1, wherein   The execution instruction unit has a function of notifying the proxy server unit of a request generation unit to which an operation content for each step is allocated when the operation content for each step is allocated to the request generation unit. The TAT measuring device according to claim 1. The request generation means
A scenario execution instruction receiving unit that receives an execution instruction of the operation content for each stage from the execution instruction unit;
A request transmission unit that transmits a request based on the operation content for each stage to the web server;
A response receiver for receiving a response from the web server to the request;
A response screen display unit for displaying the response on the screen;
The TAT measurement apparatus according to claim 1, further comprising: a response information transmission unit that determines whether or not the response is a valid response and records the determination result in the response information table. Comprising a plurality of request generating means for generating an operation by a plurality of users in the procedures pseudo, and operation contents to be performed, in advance stores the response parameter is a cookie information corresponding to each phase of the operation content There is a TAT (Turnaround Time) measuring apparatus, which measures a TAT of a web server in response to an operation by the user,
Search for available ones among the plurality of request generation means,
The vacant request generation means assigns and executes the operation contents for each stage and the response parameter,
Mediating access to the web server generated by the request generation means,
A TAT measuring method characterized in that a time from the access until a response to the response is returned is measured and used as the TAT. Comprising a plurality of request generating means for generating an operation by a plurality of users in the procedures pseudo, and operation contents to be performed, in advance stores the response parameter is a cookie information corresponding to each phase of the operation content In a TAT (Turnaround Time) measuring device, a computer provided in the TAT measuring device includes:
A procedure for searching for a free one of the plurality of request generation means;
A procedure for causing the vacant request generation means to assign and execute the operation content for each stage and the response parameter;
And instructions for arbitrating access to a window Ebusaba said request generating means caused,
A TAT measurement program for measuring a time from the access until a response to the response is returned, and executing the procedure as the TAT.

Images