Lumen

Fast HTTP load testing CLI — dynamic templates, threshold-gated CI, and load curves.

Full documentation at lmn.talek.cloud

Why Lumen

Most load testers answer "how fast is my API?" Lumen also answers "did this release break performance?" — by letting you define pass/fail thresholds and wiring the exit code into CI.

lmn run -H https://api.example.com/orders \
  --header "Authorization: Bearer ${API_TOKEN}" \
  -f lmn.yaml
# exits 0 if thresholds pass, 2 if they fail

# lmn.yaml
execution:
  request_count: 1000
  concurrency: 50

thresholds:
  - metric: error_rate
    operator: lt
    value: 0.01        # < 1% errors
  - metric: latency_p99
    operator: lt
    value: 500.0       # p99 < 500ms

Installation

cargo install lmn

Docker (zero-install):

docker run --rm ghcr.io/talek-solutions/lmn:latest run -H http://host.docker.internal:3000/api

Homebrew and pre-built binaries: see Installation docs.

Quick Start

# 100 GET requests, see latency table
lmn run -H https://httpbin.org/get

# POST with an inline JSON body
lmn run -H https://httpbin.org/post -M post -B '{"name":"alice"}'

# Run from a YAML config file
lmn run -f lmn.yaml

See the Quickstart guide for a full walkthrough.

Features

• Dynamic request bodies — per-request random data from typed JSON templates
• Threshold-gated CI — exit code 2 on p99/error-rate/throughput failures; wires into any pipeline
• Load curves — staged virtual user ramp-up with linear or step profiles
• Auth & headers — ${ENV_VAR} secret injection, .env auto-load, repeatable headers
• Response tracking — extract and aggregate fields from response bodies (e.g. API error codes)
• JSON output — machine-readable report for dashboards and CI artifacts
• Config files — full YAML config with CLI flag precedence

Observability

Stream traces to any OpenTelemetry-compatible backend:

export OTEL_EXPORTER_OTLP_ENDPOINT=http://my-collector:4318
lmn run -H https://api.example.com

Start a local Tempo + Grafana stack from lmn-cli/:

docker compose up -d
# Grafana at http://localhost:3000 → Explore → Tempo

Reference

• CLI reference — full flag and config reference
• Template placeholders — request and response template reference
• JSON output schema — machine-readable report structure

Project Structure

lmn/
├── lmn-core/     # engine, templates, HTTP, thresholds (library crate)
└── lmn-cli/      # CLI entry point, OTel setup (binary crate)

cargo build
cargo test

License

Apache-2.0 — see LICENSE.