WorkManager benchmarks

Generated 2026-11-11T04:24:52.186375+00:00 · 20 results · 0 regression(s) flagged against a rolling baseline.

Engine benchmarks

One marker worker per engine, called directly through IEngine.ProcessAsync: calls/second and per-call latency without the API, Redis or the actor turn (5 engines).

WorkManager pipeline benchmarks

Publish → API → Redis/Dapr → Runtime actor → worker → subscriber, at a fixed offered rate and at saturation: end-to-end events/s, p50/p99 and errors (3 engines).