:question: When deploying metaflow with the minima...
# ask-metaflow
a
When deploying metaflow with the minimal aws example, we ran into a situation where the terraform plan worked but the terraform apply via CI failed because there were unexpected tags. I reported the issue but I'm curious if anyone else has run into this and what, if anything, I can do to make the prod deploy via CI smoother than the dev deploy.
1
👀 1
a
that links goes to a private google doc but I think you meant this
i haven't run into this, do you use any custom tags there? there are some threads on terraform github where this happens where people try to use current timestamp in tags and some other dynamic things like that
a
we had an alternative hypothesis the issues had something to do with the interaction between our stack and this module. it sounds like that might be the case.
Could it be this part of the suggested minimal example metaflow.tf that is causing tag conflicts on apply?
Copy code
tags = {
    "managedBy" = "terraform"
  }
b
Bumping this thread! @adventurous-plastic-90730 were you able to get past this? We're currently running into this on our team
Not exactly sure what it was, but re-ran everything with the tag changed to:
Copy code
tags = {
    "managedBy" = "metaflow-terraform"
  }
and it works 🤷‍♂️
a
@bright-author-73055 we ended up forking the repo because we had other reasons to do so and managed this and some auth terraform that way.
b
@adventurous-plastic-90730 gotcha gotcha makes sense. Thanks!! 🙂
👍 1