stelligent
Mu fails when there's an existing ECR repo
Getting this error after running mu pipeline up:
The following resource(s) failed to create: [EcsRepo]. . Rollback requested by user. 11:13:52 UTC-0400 CREATE_FAILED AWS::ECR::Repository EcsRepo getmu-io already exists
Takes about 5+ mins to get to this point. Is there a way to check for a ECR repo name at the beginning and fail on the mu cli?
Wouldn't be simpler to just treat this as a warning? As long as the EcsRepo is still writable, the rest of the pipeline should be fine, since it is just a place to put new Docker images. Is there a reason it SHOULD fali completely if the repo already exists?
Justification/precedent: mkdir -p doesn't error out if the directory already exists.
@timbaileyjones not that easy. No way to tell CloudFormation to "reattach" to a resource that already exists. So, when mu goes to upsert the CFN stack for the repo, CFN will error out if it tries to create a repo that already exists with that name.
It's a bit of an edge case. The way this happens is when someone tries to delete the repo stack, gets an error that the repo isn't empty, tries to delete again and selects retain resource on the repo.
