We use this feature in Katalon to pass credentials and url’s at startup to our project Global variables and Execution profile | Katalon Docs This has stopped working in TestOps with generated commands some time ago.
I like to know if there is any other way to schedule test suite (collections) in testops with custom global variable values
At present, a TestOps schedule cannot supply arbitrary runtime -g_ / -p_ global-variable overrides to the generated KRE command. Those flags are supported by Katalon Studio/KRE, but TestOps-generated/scheduled commands do not expose a per-run parameter-injection field.
The supported TestOps-native option is to create execution profiles (for example, staging, qa, prod) and select the required profile for each suite in the suite collection, then schedule the appropriate profile/collection. Keep secrets as protected variables.
If the URL/credentials must vary for every scheduled run, run KRE from your CI scheduler instead (Jenkins, GitHub Actions, etc.) and provide -g_<name>=... for regular globals or -p_<name>=... for protected globals. You can still upload/report that run to TestOps. This keeps secrets in the CI secret store rather than in the project or schedule.
We have done it many times this way in the past using our cloud execution runners in TestOps, and only since recently this stopped working. We have ephemeral QA environments, that have unique URLs every time. We cannot schedule tests sequentially for multiple environments if we have to create or configure execution profiles for every environment. It is simply not possible.
Thank you for clarifying, you’re actualy right. For ephemeral QA environments, creating and maintaining an execution profile per URL is not a workable alternative, particularly when runs need distinct runtime values.
Because this previously worked through TestOps cloud execution runners and has since stopped, we will treat it as a regression in the generated-command/runtime-override flow.
Could you please share a redacted example (if possible) of the command that previously worked and one that now fails, along with the affected execution IDs and approximate date it changed? This will help us investigate the change and confirm the expected behavior for -g_/-p_ overrides.
profile-based approach does not scale well when every test run uses a newly created environment, because creating a separate profile for each temporary URL is impractical..
Recommended workaround
Use a CI/CD pipeline to:
Create or identify the temporary environment.
Store its URL and credentials in the CI secret store.
Run KRE with the appropriate -g_ or -p_ arguments.