Remote comms: Basic Rate Limiting

[sirtimid]

Overview

The remote communications system needs rate limiting to protect against message and connection flooding attacks.

Note: PR #714 implemented resource limits for capacity (concurrent connections, message size, queue depth). This issue now tracks the remaining rate-based limits.

What's Already Implemented (PR #714)

• ✅ 100 concurrent connections limit (maxConcurrentConnections)
• ✅ 1MB message size limit (maxMessageSizeBytes)
• ✅ Message queuing with 200 capacity (maxQueue)
• ✅ Stale peer cleanup (1 hour timeout)

Remaining Work

1. Message Rate Limiting

• Requirement: 100 messages per second per peer (sliding 1-second window)
• Behavior: Queue excess messages up to capacity, then drop with error logging
• Implementation: Track per-peer message counts with sliding windows, validate rates before sending

2. Connection Attempt Rate Limiting

• Requirement: 10 connection attempts per minute per peer
• Behavior: Reject connection attempts that exceed the rate limit
• Implementation: Log connection attempt timestamps per peer, validate before dialing

Implementation Approach

• Create a SlidingWindowRateLimiter class for tracking message rates
• Add per-peer connection attempt tracking
• Integrate rate checks into sendRemoteMessage() and dial logic
• Add error logging when messages are dropped due to rate limits

Acceptance Criteria

• Message rate limiting enforced at 100 msg/sec per peer
• Connection attempt rate limiting enforced at 10 attempts/min per peer
• Error logging when messages are dropped due to rate limits
• Unit tests verify enforcement of both rate limits
• Tests confirm proper queueing behavior when approaching limits