I’m trying to figure out how to bring in our automatic E2E test results to Zephyr Squad Server without making new Jira tickets every time. We run our tests with Cypress and Cucumber, and we’ve got junit.xml and cucumber.json files with the results.
I can update one test’s status through the Zephyr API, but it’s a pain with all the steps involved. I thought about using the “test automation task” to upload a junit.xml file, but that always makes a new Jira ticket. This messes up our ability to track a test’s performance over time.
Even though our test names in the junit.xml match those in Jira, it doesn’t help. The Zephyr docs for the Server version aren’t great, and I’m used to Xray where this process is much simpler.
Has anyone found a way to bulk import test results into Zephyr Squad Server and just add new test executions to existing tickets? Any tips would be super helpful!
hey dancingfox, ive run into this too. zephyr can be a pain sometimes lol. have u tried using the bulk update api? it lets u update multiple test results at once without creating new issues. u might need to write a script to parse ur junit/cucumber files and send the data, but it could save u tons of time in the long run
I’ve been in your shoes, and it’s definitely frustrating. One approach that worked for me was leveraging Zephyr’s REST API in combination with a custom ETL (Extract, Transform, Load) process. We set up a nightly job that pulls the test results from our CI system, transforms them to match Zephyr’s data structure, and then uses the API to update existing test cases.
The trickiest part was mapping our test identifiers to Zephyr’s test case IDs. We ended up adding a custom field to our Jira issues to store our unique test identifiers, which made the matching process much smoother. It took some initial effort to set up, but now it runs like clockwork.
This method allows us to maintain historical data without creating new issues for each test run. It’s not perfect, but it’s been a solid solution for our team. If you’re comfortable with a bit of coding, it might be worth exploring this route.
I’ve faced a similar challenge with Zephyr Squad Server, and I found a workaround that might help. Instead of relying on the built-in import functionality, I developed a custom script that interfaces directly with Zephyr’s REST API. This script parses our junit.xml and cucumber.json files, then maps the results to existing test cases in Zephyr.
The key is to maintain a mapping between your Cypress/Cucumber test identifiers and the corresponding Zephyr test case IDs. With this in place, the script can create new test executions for existing test cases without generating duplicate Jira issues.
It took some initial setup, but now our CI pipeline automatically updates test results in Zephyr after each run. This approach gives us the historical tracking we need without the clutter of new tickets. If you’re comfortable with scripting, this method could solve your problem efficiently.