Summary and Quiz

This chapter series covered Amazon DocumentDB end to end, starting from its flexible document data model and compute-storage separated architecture, through advanced topics like elastic clusters, change data capture, global replication, AI-era search capabilities, and expert-level production operations.

Document model and JSON thinking

DocumentDB stores data as flexible BSON documents rather than rigid relational rows. A single document can embed nested objects and arrays, eliminating multitable joins. Schema flexibility means each document defines its own structure, and adding or removing fields requires no table-wide migration. Data modeling is driven by application access patterns rather than normalization rules.

Cluster architecture and compute-storage separation

A DocumentDB cluster decouples compute instances from a shared cluster storage volume. The storage layer replicates data six ways across three Availability Zones and grows automatically in 10 GB increments up to 128 TiB. One primary instance handles all writes, while up to fifteen read replicas share the same storage volume without copying data. This separation means scaling compute is independent of storage capacity.

Endpoints and application connectivity

Three endpoint types abstract infrastructure changes. The cluster endpoint always resolves to the current primary for writes. The reader endpoint load-balances across available replicas for reads. Instance endpoints target specific replicas for specialized workloads like diagnostics. Applications should always use cluster and reader endpoints in ...