Quick Baseline¶

Test a server with reasonable defaults:

vex --target example.com

Runs 1000 requests with 1 worker to establish baseline latency.

Concurrency Testing¶

Measure performance under increasing load:

# Light load: 10 workers
vex --target example.com --workers 10 --requests 1000

# Medium load: 50 workers
vex --target example.com --workers 50 --requests 5000

# Heavy load: 200 workers
vex --target example.com --workers 200 --requests 10000

Time-Based Testing¶

Run for a fixed duration to find sustained throughput:

# 1 minute test
vex --target example.com --workers 100 --duration 60

# 5 minute test
vex --target example.com --workers 50 --duration 300

# Long soak test
vex --target example.com --workers 25 --duration 3600

Endpoint Testing¶

Test specific API endpoints:

# REST endpoint
vex --target api.example.com --path /v1/users --workers 50 --requests 5000

# GraphQL endpoint
vex --target api.example.com --path /graphql --workers 100 --requests 10000

# Health check endpoint
vex --target service.local --path /health --workers 10 --requests 1000

Local Service Testing¶

Test local services with custom ports:

# Local HTTP/3 service on port 8443
vex --target 127.0.0.1 --port 8443 --insecure --workers 50

# IPv6 localhost
vex --target "[::1]" --port 8443 --insecure

Debugging¶

Enable verbose mode to see request/response details:

# Small test with header output
vex --target example.com --workers 1 --requests 10 --verbose

Shows: - HTTP status and headers for each request - Latency for each request - Any errors encountered

Mixed Constraints¶

Use both request count and duration to explore different scenarios:

# Stop when 10000 requests complete OR 60 seconds elapse
vex --target example.com --workers 100 --requests 10000 --duration 60

# Will complete when:
# - All 10000 requests are sent, OR
# - 60 seconds have passed
# Whichever comes first

Check the output for "Completion reason" to see which limit was reached:

Completion reason: All 10000 requests completed

or

Completion reason: Duration limit (60s) reached

Performance Comparison¶

Compare two configurations:

# Baseline
vex --target service1.example.com --workers 50 --requests 5000

# New version
vex --target service2.example.com --workers 50 --requests 5000

Compare the latency metrics and request/sec values.

High Throughput Testing¶

Maximize throughput with many workers:

vex --target example.com \
    --workers 500 \
    --requests 100000 \
    --duration 120

Monitor: - Requests/sec value - p99 latency for tail performance - Error counts for stability

Diagnostic Run¶

Small test to verify connectivity and get baseline:

vex --target example.com \
    --workers 5 \
    --requests 100 \
    --verbose

Output shows: - If connection succeeds - Response times - Any errors - HTTP status distribution