Request rate limiting to protect PLC resources

[gijzelaerr]

Summary

Add configurable request rate limiting to prevent overwhelming PLCs with too many requests, especially older/smaller models with limited processing capacity.

Background

Siemens S7-200 and S7-300 PLCs have limited communication resources. Sending requests too fast can cause the PLC to drop connections, queue requests, or even affect scan cycle times. A built-in rate limiter would protect against this, especially in polling loops.

What needs to be done

• Implement a configurable rate limiter (max requests per second)
• Support both fixed-rate and token-bucket algorithms
• Apply rate limiting transparently to all client operations
• Add per-connection and optionally per-PLC-model rate limits
• Add configurable behavior when limit is hit: block (wait), raise exception, or drop
• Make it opt-in (disabled by default)
• Add tests
• Add documentation

API Design Ideas

client = snap7.Client(
    max_requests_per_second=10,  # 0 for unlimited (default)
    rate_limit_behavior="block",  # "block", "raise", or "drop"
)

client.connect("192.168.1.10", 0, 1)

# Operations are automatically throttled
for i in range(1000):
    data = client.db_read(1, 0, 4)  # Will not exceed 10 req/s

Notes

• Different PLC models have different limits:
  • S7-200: ~10 requests/second
  • S7-300: ~20-50 requests/second
  • S7-400: ~100+ requests/second
  • S7-1200/1500: ~100+ requests/second
• Could provide sensible defaults per PLC model (auto-detected from CPU info)
• Rate limiting should account for multi-variable operations (one multi-read = one request, not N)
• Should work with both sync and future async clients