We are currently using a self-hosted instance of Gitlab 16.1.2. We mainly use dynamic child pipeline until today where we keep getting the following error on previously working pipelines:
We are using kubernetes executor runners deployed with the gitlab-runner Helm chart. No strange logs shown within the runners.
I could not find any information on this specific undefined error…
The error says “yaml invalid”. The yaml generated is stored as an artifact, so I’d suggest finding the artifact and looking at what is being generated.
I tried to run a pipeline with the exact same content of my downloaded artifact gitlab_config_child.yaml, with a simple copy-paste. So my .gitlab-ci.yml looks like this
Problem solve after updating Gitlab Instance from 16.1.2 to 16.4.2.
For information, our Gitlab runners are still in 16.1.1. So it clearly seems to be an issue with our Gitlab Instance.
hi, guys
current we use the version v16.9.2 via docker-compose installed. and the gitlab-runner installed same with you on helm chart, but the version big than gitlab instance, and now we facing the same error as you noted, any idea on help?
thanks
Possibly, the workaround for 7.3.1 is to restart all gitlab nodes or:
Alternatively, navigating to the Admin Area → CI/CD Settings (/admin/application_settings/ci_cd), expanding the Continuous Integration and Deployment section and pressing the Save button at the end there (even without changing anything) might also resolve it.