Skip to content

Latest commit

 

History

History
43 lines (30 loc) · 2.23 KB

contribute.md

File metadata and controls

43 lines (30 loc) · 2.23 KB

[MO] Contribute by adding new tests (~10min)

  • 👨‍💻 Owner: Justine Mignot
  • ⏰ Last update date: 26th of April 2018

Steps

Add a new API

  • Create a new folder in ./src, for example: ./src/apiName3
  • Copy an index.js file like: ./src/apiName3/index.js
  • Create another folder, tests in your API folder, example: ./src/apiName3/tests
  • Check the next steps for adding tests

Add a new tested route

  • Add a new test by copying how other tests work for example: src/apiName1/tests/fetchComments.js

  • Add your test function in the API index.js, for example: src/apiName1/index.js

  • Add the test threshold in the global options

  • Add your test and its instance to index file

  • Run the test directly on the new route to get its average response time

    k6 run src/{replace_by_your_API}/tests/{replace_by_your_test}.js --vus 2 --iterations 10

    Why "vus = 2" and "iteration = 10": we use those same settings in global options to run all the tests of a full API

    docker run -v {replace_by_path_of_src_folder}:/k6 -i loadimpact/k6 run /k6/{replace_by_your_API}/tests/{replace_by_your_test}.js --vus 2 --iterations 10

  • In your test, modify field LATEST_PERFORMANCE with http_req_waiting average time.

    Request is cut in several steps : http_req_connecting, http_req_sending, http_req_waiting, http_req_receiving,... The time that matters, on which we can act is http_req_waiting time because it represents the server processing your request. Hence the performance of the code, and not the network.

  • Depending on the result:

    • If the performance is above the standard time you fixed (we chose 500ms): Andon your architect and/or PO. Depending on the gravity, either find time to investigate asap, or create a performance issue ticket in order to fix it next time.
    • If the performance is acceptable: Write in your new test file the found average performance

  • Don't forget to version the changes in your performance monitoring repo

👉 Next steps: Follow Run K6 on an API