Amazon Aurora DSQL

A dated reading of what is claimed, reported, and independently verified in the current evidence.

Current reading

Readiness band and full integer triple

AI-assisted assembly · derived results

Readiness band

Watch

Primary summary from verified readiness

Evidence strength Strong

Decision

Agent-assisted assessment · independently reviewed · 2026-10-21

Consider production use for compatible, purpose-built transactional applications after workload testing.

Why: Sustained production experience and independent benchmarks show practical viability, while SQL compatibility, write contention and regional latency constrain fit.
Next: Pilot the intended Region topology with representative data; test retries, migrations, recovery, transaction correctness and DPU costs before committing.

Constraints

PostgreSQL migrations can require application changes; SAVEPOINT, nested transactions and some column alterations remain unsupported.
Hot-row writes can exhaust optimistic concurrency retries and worsen response times.
Multi-Region commits add write latency; measure the intended Region pair against application latency targets.
Write transactions are limited to 3,000 mutated rows, 10 MiB of modified data and five minutes.
Connections expire after 60 minutes, and new connections are capped at 100 per second; connection reuse and IAM integration require care.
ORM upgrades need compatibility testing: a September Django canary failed while supported pinned releases remained unaffected.
DPU costs depend on workload and query plans; a 2025 TPC-B-like comparison favored Aurora Serverless v2 on price and throughput.

Evidence summary

AI-assisted assembly

Counts and dates only. Raw signals, private excerpts, trust records, and internal corpus material are not published here.

Publication record

Revision 5