Briefing
gateway 2.3.0 went out ten minutes ago with a refactored manifest.
Both pods started straight away and neither has restarted once, yet neither has ever reported Ready, and the rollout has now blown through its progress deadline.
This is the gateway's first deploy to this cluster, so there is no older ReplicaSet to fall back on, and partner API calls are failing.
Sam thinks 2.3.0 is hanging on a new startup dependency and wants to roll back to 2.2.
The fix is yours to find: hints, the recovery checks and the debrief unlock inside the incident.