I'm looking at your docs, and the RUN REPEATABLE command seems really powerful. But if the state is broken after a run, like if you have some pods stuck at Terminating, how would you recover things?
Another question I have is how would you handle state that sometimes needs to update and sometimes doesn't? For example, it would be ideal to have a staging database that can keep having migrations and data added to it when new features are added, but we only want to checkpoint the changes to it from testing when the PR is actually merged.
Another question I have is how would you handle state that sometimes needs to update and sometimes doesn't? For example, it would be ideal to have a staging database that can keep having migrations and data added to it when new features are added, but we only want to checkpoint the changes to it from testing when the PR is actually merged.