torks 0.9.5

Fast HTTP load testing CLI built in Rust
torks-0.9.5 is not a library.

Torks is a lightweight, high-performance HTTP load testing tool with a real-time TUI dashboard, interactive mode, and CI-friendly output. Built with Rust and Tokio for maximum throughput.

Features

  • Real-time TUI dashboard with RPS & latency charts
  • Interactive quick mode — just run torks quick and follow the prompts
  • Dual HTTP engine: reqwest (full-featured) or hyper (raw speed)
  • Configurable worker threads via --threads
  • TOML-based scenario files with multi-step flows
  • HDR Histogram for accurate percentile tracking
  • Variable extraction from JSON responses
  • Weighted scenario selection
  • Ramp-up scheduling
  • JSON & CSV report export
  • CI/CD threshold checks (p99, error rate, min RPS)
  • Quiet mode for headless environments

Installation

cargo install torks

Quick Start

Interactive Mode

Run torks quick with no arguments to launch the interactive wizard:

torks quick

You'll be guided through:

  • HTTP engine selection (reqwest or hyper)
  • Worker threads (auto or custom count)
  • Target URL & path
  • HTTP method (arrow-key selection)
  • Headers (add as many as needed)
  • Request body (for POST/PUT/PATCH)
  • Concurrent users, ramp-up, duration, timeout
  • Summary & confirmation before launch

One-liner Mode

# GET request with 50 users for 30 seconds
torks quick https://api.example.com/health -u 50 -d 30s

# POST with 100 users, 5s ramp-up
torks quick https://api.example.com/data -m POST -u 100 -d 1m -r 5s

# Quiet mode (no TUI, for CI)
torks quick https://api.example.com/health -u 10 -d 10s -q

# Use hyper engine for raw speed
torks quick https://api.example.com/health -u 100 -d 30s --engine hyper

# Control worker threads (default: CPU core count)
torks --threads 4 quick https://api.example.com/health -u 200 -d 1m

Scenario File

# Run a scenario file
torks run scenario.toml

# With JSON and CSV reports
torks run scenario.toml --json report.json --csv report.csv

# Quiet mode
torks run scenario.toml --quiet

Validate Config

torks validate scenario.toml

Commands

Command Description
torks Show banner with usage info
torks quick Interactive load test wizard
torks quick <URL> Quick load test with flags
torks run <FILE> Run a TOML scenario file
torks validate <FILE> Validate a scenario file without running
torks --help Show all options

Global Flags

Flag Default Description
--threads CPU core count Number of tokio worker threads

torks quick Flags

Flag Short Default Description
--users -u 10 Number of concurrent users
--duration -d 10s Test duration (30s, 5m, 1h)
--method -m GET HTTP method
--ramp-up -r 0s Ramp-up time
--engine reqwest HTTP engine: reqwest (full-featured) or hyper (raw speed)
--quiet -q false Disable TUI dashboard

torks run Flags

Flag Short Default Description
--json Output JSON report to file
--csv Output CSV report to file
--quiet -q false Disable TUI dashboard

Scenario File Format

Scenario files use TOML format:

[target]
base_url = "https://api.example.com"
timeout_ms = 30000

[target.headers]
Authorization = "Bearer your-token"
Content-Type = "application/json"

[[scenarios]]
name = "browse-api"
weight = 70

[[scenarios.steps]]
method = "GET"
path = "/users"

[[scenarios.steps]]
method = "GET"
path = "/users/{{user_id}}/posts"

[[scenarios]]
name = "create-post"
weight = 30

[[scenarios.steps]]
method = "GET"
path = "/users"
expect_status = 200

[scenarios.steps.extract]
user_id = "$.data[0].id"

[[scenarios.steps]]
method = "POST"
path = "/users/{{user_id}}/posts"
body = '{"title": "Load Test", "content": "Hello"}'

[scenarios.steps.headers]
Content-Type = "application/json"

[load]
users = 100
ramp_up = "10s"
duration = "2m"

[thresholds]
p99_ms = 500
error_rate = 5.0
min_rps = 100.0

Config Reference

[target]

Field Required Default Description
base_url Yes Base URL for all requests
timeout_ms No 30000 Request timeout in milliseconds
headers No {} Default headers for all requests

[[scenarios]]

Field Required Default Description
name Yes Scenario name
weight No 100 Weight for random selection
steps Yes List of HTTP steps

[[scenarios.steps]]

Field Required Default Description
method Yes HTTP method (GET, POST, PUT, DELETE, PATCH, HEAD, OPTIONS)
path Yes URL path (appended to base_url)
headers No {} Step-specific headers
body No Request body
extract No {} Extract variables from JSON response
expect_status No < 400 Expected HTTP status code

[load]

Field Required Default Description
users Yes Number of concurrent virtual users
duration Yes Test duration (30s, 5m, 1h)
ramp_up No 0s Time to ramp up to full users

[thresholds]

Field Required Description
p99_ms No Max p99 latency (ms) — exits with code 1 if exceeded
error_rate No Max error rate (%) — exits with code 1 if exceeded
min_rps No Min requests/sec — exits with code 1 if below

Variable Extraction

Chain requests by extracting values from JSON responses:

[[scenarios.steps]]
method = "POST"
path = "/auth/login"
body = '{"email": "test@example.com", "password": "secret"}'

[scenarios.steps.extract]
token = "$.data.access_token"

[[scenarios.steps]]
method = "GET"
path = "/api/profile"

[scenarios.steps.headers]
Authorization = "Bearer {{token}}"

HTTP Engines

Torks ships with two HTTP engines:

Engine Flag Best For
reqwest (default) --engine reqwest Full-featured tests: cookies, JSON, TLS, headers, multi-step scenarios
hyper --engine hyper Raw throughput benchmarks: minimal overhead, maximum RPS 
# Default engine (reqwest) — full feature support
torks quick https://api.example.com/health -u 100 -d 30s

# Hyper engine — raw speed, ~2x RPS
torks quick https://api.example.com/health -u 100 -d 30s --engine hyper

CI/CD Integration

Use --quiet mode and thresholds for CI pipelines:

# GitHub Actions example
- name: Load Test
  run: |
    cargo install torks
    torks run load-test.toml --quiet --json results.json

Torks exits with code 1 if any threshold is exceeded, making it easy to gate deployments.

License

Apache-2.0

Author

Built by Armonika