A gaming studio needs sub-10ms reads for player profiles with millions of requests per second, flexible JSON attributes per game, global distribution, and strong consistency only for inventory trades while profiles use eventual consistency. Which database service best fits these access and scalability requirements?
Select an answer to reveal the explanation.
Short Explanation
Millions of tiny profile lookups need speed, not joins. DynamoDB scales sideways effortlessly and lets you dial consistency per call, with tunable consistency strong only where money moves. It's a locker room built for lightning-fast key lookups.
Full Explanation
DynamoDB provides single-digit millisecond key-value access at massive throughput, schemaless JSON-like items for evolving game attributes, global tables for multi-region distribution, and per-request consistency control allowing strongly consistent inventory trades and eventually consistent profiles. Aurora PostgreSQL offers relational JSONB and read scaling but cannot sustain millions of simple key lookups at sub-10ms globally without complexity and consistency is instance-level, not per-request. RDS MySQL with Multi-AZ and replicas improves availability and read offload but remains relational with higher latency and limited elastic write scaling for this pattern. Redshift is an OLAP warehouse for analytical scans, not low-latency transactional lookups. Exam caveat: choose DynamoDB for massive key-based scale with mixed consistency, relational for complex joins and transactions. Practical check: model access patterns around partition keys and verify DAX hit rates for repeated profile reads.