You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Calling this a feat because I believe technically we are updating the functionality of gc.
Previously we were waiting for stacks in `REVIEW_IN_PROGRESS` to land, because that is the one CFN state that you cannot retrieve a template for (because it doesn't exist yet). However in environments where we are constantly deploying new stacks (like our test environments), we may never get to a state in the allotted time where no stacks are `REVIEW_IN_PROGRESS`.
Instead, we are going to ignore `REVIEW_IN_PROGRESS` stacks. This will introduce a subtle race condition where a previously isolated asset becomes in-use by the `REVIEW_IN_PROGRESS` stack before it turns into a `CREATE_IN_PROGRESS` stack and we can reference its stack. If garbage collection happens to come across the isolated asset while the stack is `REVIEW_IN_PROGRESS` (aka before it is `CREATE_IN_PROGRESS` but after CDK has verified that the assets exist) we will garbage collect the asset. However, we don't expect this to become a big problem in practice.
----
*By submitting this pull request, I confirm that my contribution is made under the terms of the Apache-2.0 license*
0 commit comments