Performance

For operators sizing a deployment, and for anyone evaluating Dwarf. Every number here was measured, and the configuration that produced it is stated alongside.

• Cloud Benchmarks — Dwarf measured against managed cloud PostgreSQL across a real network hop. How much one database serves, what the engine costs on top of it, what adding shards and engine replicas does, long tasks, failover, payload and state size. The recipe built from them is Sizing a Deployment.
• A Day of Varied Load — one database through 4.5 hours of mixed flows, bursts above its capacity, a rolling restart of the engines and a restart of the database itself.
• Sizing & Tuning — how the engine sets connection budgets and worker counts, and the measurements behind them, for operators who want to understand or override them.
• Dialect Benchmarks — in-repo throughput and latency benchmarks per SQL dialect, run on a single machine. Use them to compare dialects on your own hardware, not to size a deployment.

Next: Cloud Benchmarks.