HydraDB: a Rust graph database that lives entirely in S3, with nothing on disk
thisdudelikesAI · x · 2026-09-08
HydraDB, a Rust-based open-source graph database (AGPL), fully separates compute and storage: an S3-compatible bucket holds graph records, WALs, manifests, and immutable traversal indexes, while query nodes keep only disposable memory and SSD cache, rebuilding themselves from durable state.
Key technical details:
- Writer conflicts are handled with object-store CAS leases per cell plus SlateDB writer epochs to fence stale writers at the storage layer
- Reads run against one pinned SlateDB snapshot; when an index is missing or behind, the visible WAL tail is replayed on top so readers stay correct
- Supports an OpenCypher subset (variable-length paths, OPTIONAL MATCH, UNION, aggregation), is compatible with Neo4j drivers over Bolt 5.x, and uses SuiteSparse GraphBLAS for traversal
- Causal and strong read modes, Helm chart, Prometheus metrics, OpenTelemetry tracing
- Unusually thorough for a 13-star project: a Jepsen consistency report, Quint formal verification evidence, and a correctness casebook with regression tests
More from Infra
- Yacine urges building sovereign AI infrastructure or losing privacy and IP — yacineMTB · 2026-09-08
- Crusoe raises $3B at $30B valuation as it builds Stargate, delivered 200MW in 11 months — 快鲤鱼 · 2026-09-08
- CPU-Only LLM Tests: 35B MoE at Q2 Beats a 2B Model Despite Half the Speed — ML-Future · 2026-09-08
- Running Qwen 27B and Gemma 31B locally on one RTX 4090: quantization and context tradeoffs — MooseEfficient2151 · 2026-09-08
- Could distributed iPhones form an inference network? A 2AM open question — gajesh · 2026-09-08
- Candidate depth 10→500 lifts score by just 0.01: Qdrant on diagnosing before tuning vector search — qdrant_engine · 2026-09-08