Create alert /warning mechanism and process for unclean TF plan on gprd / ops
From DNA discussion, we would like to have better process around an andon cord for unclean TF plans for gprd and ops.
Current ideas:
- Run a TF plan in ci for gprd and ops every 3(6)0 minutes
- Capture the footer and if above a certain threshold, create an issue? with that data. We want to understand the trends.
- Could we use the change lock idea and extend that here? If 2 above is failing, we change lock tf gprd?
- Side discussion around TF enterprise or Atlantis? are we re-inventing the wheel when we don't need to?
Adding a due date to end discussion and implement something. Due date for now is end of discussion and choose what to implement as first increment toward improving blockages on TF.
cc @craig, @craigf, @cmcfarland, @bjk-gitlab, @T4cC0re, @ggillies for starters of who has spoken up on this topic