Why it matters
DynamoDB removes the ops burden of running Cassandra or HBase. Trade-off is AWS lock-in and pricing that grows with usage. For most workloads, the operational simplicity is worth it.
The architecture
A DynamoDB table has a partition key (hash key) that determines physical storage location. Optional sort key enables efficient range queries within a partition. Together they form the primary key.
Global Secondary Indexes (GSI) enable queries on non-key attributes. Local Secondary Indexes (LSI) enable additional sort keys on the same partition.
How it works end to end
Capacity modes: on-demand (pay per request, no capacity planning) or provisioned (set RCUs/WCUs, cheaper at scale). Auto-scaling adjusts provisioned capacity based on load.
DynamoDB Streams captures every change to the table. Lambda can subscribe for event-driven processing.
Global Tables replicate across regions for multi-region access. All writes eventually consistent across regions.