Throttle message stream to target rate by selective forwarding

Use This When

• Protecting downstream components or UIs from overwhelming message rates
• Smoothing bursty data streams to consistent rate for recording or display
• Reducing processing load by sampling high-frequency sensor or video streams
• Implementing rate limiting for API endpoints or network-constrained outputs

What It Does

• Estimates current message rate using exponential moving average or token bucket algorithm
• Selectively forwards or drops messages to converge to target messages per second
• High rate mode (≥1 msg/s): uses EMA to minimize error from target rate
• Low rate mode (<1 msg/s): uses token bucket for precise fractional rate limiting

Works Best With

• High FPS video → this component → visualization at display-appropriate frame rate
• Sensor telemetry → this component → database storage at reduced sampling rate
• Integration with output-browser-video to prevent browser rendering overload
• Feeding network sinks that have bandwidth or rate constraints

Caveats

• Drops messages to achieve rate; data is lost not delayed; use queueing if buffering needed
• Time constant affects smoothness vs responsiveness tradeoff; tune based on burst patterns
• Rate estimation reactive not predictive; brief bursts may exceed target before adaptation
• No message prioritization; drops are based solely on timing not content importance