Lux Docs

Netrunner includes built-in benchmarking tools for measuring transaction throughput, block production rate, and network latency.

TPS Measurement

Measure transactions per second on a running network:

# Send 10,000 transactions and measure throughput
netrunner bench --tps --tx-count 10000

# Sustained load for 60 seconds
netrunner bench --tps --duration 60s --rate 500

Latency Profiling

Measure transaction confirmation latency at various percentiles:

# Measure p50, p95, p99 latency
netrunner bench --latency --percentiles 50,95,99

# With specific transaction count
netrunner bench --latency --tx-count 5000

Block Production Rate

Measure block production and finality:

# Measure blocks per second
netrunner bench --block-rate --duration 120s

Resource Monitoring

Track CPU, memory, and disk I/O during benchmarks:

# Enable resource monitoring
netrunner bench --tps --monitor-resources --output results.csv

Benchmark Parameters

Parameter	Flag	Default	Description
Transaction count	`--tx-count`	10000	Total transactions to send
Duration	`--duration`	60s	Test duration
Rate	`--rate`	unlimited	Transactions per second target
Output format	`--output`	stdout	Output file (CSV or JSON)
Warmup	`--warmup`	5s	Warmup period before measurement

Interpreting Results

TPS Report:
  Total TX:        10000
  Duration:        23.4s
  Throughput:      427.3 tx/s
  Avg Latency:     12.3ms
  P50 Latency:     10.1ms
  P95 Latency:     25.7ms
  P99 Latency:     42.8ms
  Failed TX:       0

Optimization Tips

Use in-memory database for maximum throughput
Reduce consensus sample size for faster finality
Increase node count only when testing scalability
Disable logging during benchmarks (--log-level=OFF)

Benchmarks