LoadRunner Skills Competition Welcome to your LoadRunner Skills Competition Customer says they need 95th percentile response time value for all transactions. Your scenario has 10 vusers which you know will generate at least 90 observations for each transaction. Your colleague says you must change your scenario to ensure 95th percentile can be calculated by LoadRunner. You decide to: Double the number of vusers to 20 Increase the number of vusers to 100 Leave the scenario unchanged None For LoadRunner Agent to be running in process mode on a Windows machine, what must be true? A user account must be logged in LoadRunner Agent must be installed as a service No one should be logged in to the Windows machine None of the above None You have 10 observations for response times (all in seconds) - 1,2,3,4,5,6,7,8,9,10. The correct values for average and 90th percentile are: 5.5 and 9 5 and 9 5.5 and 9.5 A minimum of 100 observations are needed to calculate 90th percentile None Your test scenario has 200 vusers. Beyond 50 vusers, you notice that response times stayed the same but Total transactions per second and Throughput (bytes) went down. You decide to: Tell the development team there is a bottleneck in their application Investigate your load generators Modify the scenario to add more vusers All of the above None Your script has a LoadRunner parameter called ticket_amount. In order to print its value for debugging purpose the right code is: lr_output_message("Ticket amount: %s", lr_eval_string("{ticket_amount}")); lr_output_message("Ticket amount: %d", ticket_amount); lr_output_message("Ticket amount: %s", ticket_amount); lr_output_message("Ticket amount: %d", lr_eval_string("{ticket_amount}")); None A character array called username can be declared in C as follows: char username[] = "admin"; string username = "admin"; ansi_string username = "admin"; None of the above None Run-time settings in the Controller allow you to select either thread or process mode for running your test. You prefer: Process mode because it uses fewer processes causing less stress on the load generator Thread mode because threads use fewer resources on the load generator Process mode because it creates more load on the system under test Thread mode because it creates more load on the system under test None The following declaration in a vuser script: Declares an integer whose value is 8 Declares a macro whose value is 8 Will be handled by the compiler not the preprocessor None of the above None You can create blocks in run-logic: Only in the Controller Only in VuGen In both Controller and VuGen They are only available in TruClient scripts None Receive Request Timeout value should be: Less than or equal to Step Download Timeout Higher than Step Download Timeout Always equal to Step Download Timeout None of the above None Customer's website allows keyword search (e.g. "Orville Wright Airplane") of all patent documents dating back to the 1950s. You have parameterized the search terms. Due to large expected load the number of vusers required in the scenario is ~1000. The system is known to cache the search results. The preferred data selection should be set to: Sequential Random Unique Does not matter None Sequential data selection setting for a parameter should be used: For data validation only For running a formal baseline test through the controller For running a regression test through the controller When at least 10 vusers are setup in the Controller scenario None lr_param_sprintf can be used to create a new parameter from a string. True False None IP spoofing configuration on a Windows Load Generator: Can use fake IP addresses such as 10.1.1.2, 10.1.1.3 etc. Is unnecessary because every Load Generator has a unique IP address Can use DHCP configuration to generate as many IP addresses as needed Is essential to testing load balancer behavior None To prevent making requests during replay to hosts in a web HTTP/HTTP script that are of no interest to you (say external sites outside of your organization), you can: Prevent them from being recorded in the first place by using correct recording options Comment or delete them from the script after recording Use download filters in the run-time settings All of the above None VTS can be used: As a source of parameter values for use as test data As a way to record data captured during a test As both a source of test data as well as a way to record data captured during a test None of the above None The picture below shows Analysis results for a web application. Average hits per second shown is the same as Average number of HTTP requests made every second Average number of transactional hits made per second Average number of HTTP requests made per second to retrieve HTTP resources (such as images, javascript, stylesheets, etc.) Average number of HTTP requests made per second excluding HTML pages None Your customer has strict Response Time SLAs. You ran a test but forgot to add SLAs in the scenario before running it. You can: Add the SLAs to the scenario and rerun the test Add the SLAs to Analysis and analyze the test results Never done SLAs before None A test scenario has 50 vusers. Think time is set to vary randomly between 50% to 150% of the recorded value. Pacing is set to 10 seconds. The test script has 5 transactions. The 95% response time SLA is 5 seconds. The test results indicate the SLAs were met. Customer asks if the system can support 50 simultaneous users in production. You say: Yes No Depends None The vuser_init and vuser_end actions: Are executed once per iteration Can be safely deleted Are executed only once even for multiple iterations Do not apply to Web/HTTP scripts None I certify that all information provided is accurate and that I, the person identified above, have completed this exam, in good faith, without any outside help. A client site has a homepage with several images, CSS and JS resources needed to render the page and wants to evaluate the performance for first-time users. The Web/HTTP script, with default Run-Time Settings, has a web_custom_request step with mode=HTML. Select the most efficient strategy to download all the resources in order to simulate a browser page load: Add a series of steps after the web_custom_request to get all the resources Add a series of steps after the web_custom_request in a concurrent block Either a or b None of the above None A system is known to have 100 concurrent users at peak load. In a scenario designed to simulate this load you would need 100 Vusers. True False None Time's up Posted in .
