My team decided to migrate a core service to a new database without a phased rollout. I thought the risk was too high and said so in the design review — I put together a doc outlining three failure scenarios and presented it to the group. The team heard me out but decided to proceed on the original timeline. I respected the decision and supported the launch. It actually went smoothly, so in hindsight they were probably right. I focused on making sure my part of the migration was solid.
We were debating a database migration with no phased rollout. I raised the risk formally — I modeled three failure scenarios in a written doc showing potential impact of up to four hours of downtime affecting roughly 200,000 daily active users. The team weighed it and decided to proceed on the original timeline. Once that decision was made, I owned it completely. I drafted the rollback runbook myself, set up enhanced monitoring dashboards, and volunteered for on-call the night of the launch. My concern was on record. My job from that point was to make their decision succeed — and it did, with zero incidents.