CODEWITHSUNDEEP
automationloadrunner
Siva
Siva

Reputation: 3

Automating LoadRunner execution

I have a requirement to automate the Loadrunner execution by doing some pre-checks. Steps involved would be checking for a new application build and Loadrunner/Performance center should start the load test.

Upvotes: 0

Views: 621

Answers (3)

tedyyu
tedyyu

Reputation: 607

For LoadRunner Enterprise, it supports integrations with a few popular CI systems to fulfill your requirment, like Jenkins, Azure Devops, etc. Learn more from the online help.

Upvotes: 0

Subhendu Mangaraj
Subhendu Mangaraj

Reputation: 1

After installing Load runner, you will have to create new script by selecting the correct protocol for communication between Server and the client in LoadRunner and then start recording the scripts. Next step would be to run them after parameterization required data. Please find more details at http://performancetestworld.com/JSPFiles/LoadRunnerFirstScript.jsp

Upvotes: 0

James Pulley
James Pulley

Reputation: 5682

See HP Documentation on Jekins and web services automation framework.

See command line options for mdrv and/or wlrun

Ask difficult questions. What is the end goal of the automation of the execution? How will the developers be "controlled" to not introduce structural changes to the conversation which would require business process updates to each build? How will you add actionable intelligence to the analysis of the tests?

You may be better off coupling small performance unit tests at the dev stage to ensure that code is performant at the unit and component assembly stages where the cost to fix is very small, combining this with performance checks at the functional stage (see developer tools and rules related to performance) as well as tracking response times for a single user on all business processes and then a daily/nightly execution of a business level performance test.

With performance comes ownership. If you have a working test and a developer changes something structural in their code (removes/adds web page elements, form fields, dynamic elements) and this information is not broadcast forward to test to address the changes in the test definition before the test fails, then that developer should be called out for breaking the build. The ad hominem of "Well, it didn't change the screen so it shouldn't matter...." illustrates an architectural naïveté on which OSI layer performance testing tools operate at.

Upvotes: 1

Related Questions