A team picks a rolling deployment for its API because "it's simpler." An incident review later asks why rollback took eleven minutes. What actually made it slow, and what would blue-green have done differently?

A rolling deploy replaces instances one at a time — start a new one, wait for it to be healthy, stop an old one, repeat — so a rollback is just another roll, in the other direction, and it takes as long as the original roll did: eleven minutes to roll forward is roughly eleven minutes to roll back. Blue-green keeps two complete environments and switches between them with one operation — flipping a router or a symlink — so rollback is that same switch run backward, and it's seconds, not minutes, because the old environment never stopped running; it was sitting idle the whole time, still warm. The trade blue-green makes for that speed is capacity: it needs roughly double the resources during the deploy (or all the time, if the idle environment is kept warm), which a rolling deploy never needs because it only ever runs the old and new count, never both in full.