CONCEPT Cited by 1 source
Pluggable storage backend¶
A pluggable storage backend is an architecture where the application logic layer (query engine, computation engine) is decoupled from the persistence layer via a well-defined storage interface. This allows operators to swap storage engines without changing the application layer.
Why it matters¶
- Leverage existing infrastructure: Use a proven, managed store (e.g., DynamoDB, Cassandra) rather than building custom distributed storage.
- Independent evolution: Iterate on graph/query logic without reinventing the wheel on distributed storage operations.
- Future-proofing: Migrate to a different backend as requirements change (latency, cost, operational model).
Seen in¶
- sources/2026-05-19-airbnb-scaling-identity-graph-unified-knowledge-graph-infrastructure — JanusGraph's pluggable storage backend is cited as a "unique advantage" enabling Airbnb to use DynamoDB for persistence while maintaining full control over graph logic. "We were able to leverage the scalability and reliability of AWS DynamoDB for the persistence layer while maintaining full control over the graph logic layer." They can evolve the storage layer independently as their internal persistence platform matures.
Related¶
- systems/janusgraph — canonical example (supports DynamoDB, Cassandra, HBase)
- systems/dynamodb — Airbnb's chosen storage backend for JanusGraph
- concepts/control-plane-data-plane-separation — related architectural separation principle