Open nwb-ilan opened 4 years ago
The cleanup here is necessary - it's possible to encounter errors while trying to deploy a stack (invalid serverless.yml, too many stacks/buckets on the account, etc)
It's better to try the cleanup and fail if the stack wasn't deployed than not perform any cleanup.
I do believe we should rethink the retries though
Maybe we should make sure the reason why stack deploy failed is properly reflected, otherwise the stack will be deleted and it will be difficult to find out why
We don't deal with deploying the stack ourselves, it's sls's job to show why failures happened
Before this fix, when deploying
IncreasingLoadLvl1
for example, and maven is not installed on my machine, I experience 10 retry attempts to remove the stack even though it failed to deploy.