advancedSystem Design132 of 132
How do you recover when the entire service becomes unavailable?
Tests whether you have a concrete recovery sequence -- failover to a healthy replica/region, drain and restart affected instances, and replay any buffered/queued work -- rather than a vague 'restart it and see.'
Ready to design this system end to end?
Generate a complete, structured system design answer — requirements, capacity estimation, API design, architecture, database choice, scaling, caching, fault tolerance, security, trade-offs, and more, walked through the way a strong senior engineer would in a real interview.
Sign in to generate a response