All glory and honor to God יהוה in the name of Yeshua the Messiah (Jesus Christ).

Rullst Benchmark Hub

Historical Criterion measurements for selected, named operations. Compare runs from the same CI environment; these are regression signals, not application throughput, capacity, or universal-framework rankings.

⚡ Continuous Benchmarking

Pushes to the active main branch and the weekly schedule run Criterion suites. Alerts are informational and do not currently block a merge.

📊 Selected Request Operations

The suite times in-process router calls, HTML rendering, and one WAF middleware path. It does not measure network requests per second.

🧠 CPU Microbenchmarks

Criterion reports operation timing for the exact benchmark code. This dashboard does not measure heap allocation or process memory.

Loading benchmark data...

Evidence fit matters more than one chart per crate

A benchmark is added only when it has a stable operation, input and interpretation. Other crates currently rely on evidence that better matches their risk; this table is not a maturity ranking.

SurfaceCurrent primary evidenceUseful future measurement
Messaging & MailDeterministic broker/queue, retry, restart and provider protocol contracts.Bounded local publish/claim throughput and queue-depth latency under fixed contention.
IoTProtocol vectors, signed OTA invariants, no_std target builds and simulator boundaries.Encoder cost on named host and embedded targets; real device evidence remains separate.
Macros & CLICompile-fail UI tests, generated-project compilation and runtime scaffold contracts.Cold/warm compile and scaffold time on pinned toolchains without calling it runtime speed.
Studio & NexusAuthenticated/local-access contracts and real telemetry/data integration tests.Whole-request latency and memory for a fixed dataset and page workload.

All glory and honor to God יהוה in the name of Yeshua the Messiah (Jesus Christ).