superdock

I just ran into a similar pattern on a KnownHost route to FRA last week and the carrier never owned up to it. Also 445 days is a hell of a necro, sorry.

Edge case but not really. Pop that refund to a custodial origin and you get cache hit on support tickets. Miss on self-custody and its 12 before anyone helps. Ive seen 34 on their response time. No units needed, you feel it.

The only thing that works for me is making documentation a prerequisite for change. Ansible playbook won't run without a --tags that includes a ticket number, and the ticket template forces three lines of "what did you change and why." It's annoying but it's the only enforcement that stuck.

This.