"My repository, where I post anything that I love, I learned and etc."

18 September 2020

What QA do at the start of sprint?

 Assalamualaikum,


Normally based on my previous experience, usually at the start of the sprint, we as QA, will read thoroughly the user stories first and understand what is the expected result / outcome of the user stories. If the user stories is quite a lot, we usually divide the user stories equally between the QA team members. So after dividing the user stories to respective QA, we will create manual test cases accordingly for each user story.


One user story may have 1 to 20 or more manual test cases, depending of the complexity and test scenarios that we can test for that user story. Once the manual test cases completed, we will do peer review of the manual test cases created. Sometimes the QA Lead helps in reviewing, but if the QA Lead didn't have time to review it, we normally exchange the test cases and do peer review. 


Once the story ready to be tested, then the manual QA will start to test and execute test cases that had been written in Zephyr.


As for the automation side, I also do some sort of preparation also. Since I'm the only automation QA, I need to go through all the user story and understand what is the goals or expected result for features mentioned in all user stories. I will start drafting my test script. Such as, writing or list down scenarios that I want to handle for automation. Since I'm using Protractor with Cucumber, I usually drafting the script by writing all the Gherkin syntax first, on how the process flow would like. Example of scenario:


- User enter invalid username but correct password

- User enter valid username but invalid password

- User reset both username and password field and try to click Login button


and so on...

Once the developer already come out with UI, but the function or features not really working yet, I usually inspect and get the locators of the element first in Page Object file. Once done, I'll complete writing basic test code first (the one that do normal things like sendKeys / click / scroll etc).


And once the feature finally working, I'll complete the missing codes and try running the script and fixing script if there are errors or change of element locators.





Things I get from reading articles #1

 Assalamualaikum,


Aku tahu tajuk post agak pelik tapi lantaklah. Post ni more kepada aku summarisekan / listkan input setelah aku baca beberapa artikel, so that aku boleh dapat quick info / overview tentang artikel yang aku dah baca.


1. Automate click through testing with Puppeteer 

'Puppeteer basically only runs with Chrome'.


2. E2E Testing with TestCafe and Cucumber.js

TestCafe is a Node.js tool for automated e2e web testing. TestCafe can be run on multiple browsers / platforms and can be integrated with CI pipelines (Ex : Jenkins).


TestCafe doesn't use web driver, thus, no need to upgrade we drivers for multiple browsers. Only need to update the browser.


Cucumber is basically a test runner. Cucumber use regex to map certain feature step to the written test code. Each feature step will then have failed, passed, pending or undefined flag set.


'Jenkins is not triggered by a Git Hook, instead Jenkins checks for new commit on GitHub.'