Is there an existing issue that is already proposing this?
[X] I have searched the existing issues
Describe the problem imposed by not having this feature
Sometimes we merge a PR without waiting for its CircleCI workflow to complete. When we merge, we delete the source branch. Then the subsequent job in that workflow fails on the git-shallow-clone/checkout step because the branch doesn't exist...
We would like to have a solution to prevent such workflows from spoiling our CI statistics.
Describe the solution you'd like
I propose to add an allow_fail option, set to false by default. If set to true, then the whole step would exit with code 0 even if anything would go wrong in it.
We would set it for our non-critical jobs, such as sending test coverage info to our coverage server.
Is there an existing issue that is already proposing this?
Describe the problem imposed by not having this feature
Sometimes we merge a PR without waiting for its CircleCI workflow to complete. When we merge, we delete the source branch. Then the subsequent job in that workflow fails on the
git-shallow-clone/checkoutstep because the branch doesn't exist...We would like to have a solution to prevent such workflows from spoiling our CI statistics.
Describe the solution you'd like
I propose to add an
allow_failoption, set tofalseby default. If set totrue, then the whole step would exit with code 0 even if anything would go wrong in it.We would set it for our non-critical jobs, such as sending test coverage info to our coverage server.