BestHub
Sign in
Tagged articles

adaptive throttling

4 articles · Page 1 of 1
Random Bulletin
Random Bulletin
Jul 3, 2026 · Operations

From Coarse to Fine-Grained: Traffic Shaping Strategies for Million‑QPS Queues

A bulk coupon‑sending job overwhelmed a million‑QPS system, revealing that message queues only buffer but do not shape traffic; the article dissects three failure points, compares leaky‑bucket and token‑bucket rate limiters, evaluates placement on producer, broker or consumer, and progresses from static limits to adaptive shaping with handling of throttled messages.

Rate Limitingadaptive throttlingleaky-bucket
0 likes · 18 min read
From Coarse to Fine-Grained: Traffic Shaping Strategies for Million‑QPS Queues
Random Bulletin
Random Bulletin
Jul 2, 2026 · Operations

From Zero to Control: Implementing Traffic Shaping for 10 Million QPS Systems

A bulk coupon‑sending job overwhelmed a 10 M‑QPS system, revealing that message queues only buffer traffic; the article walks through why rate‑limiting (leaky vs token bucket) must be added at the producer, broker, or consumer, evolves from static thresholds to adaptive shaping, and discusses how to handle throttled messages.

Rate Limitingadaptive throttlingleaky-bucket
0 likes · 18 min read
From Zero to Control: Implementing Traffic Shaping for 10 Million QPS Systems
Tech Freedom Circle
Tech Freedom Circle
Aug 15, 2025 · Backend Development

Calculating a 100k QPS Rate‑Limiting Threshold: Methods and Best Practices

This article explains how to determine a 100 000‑QPS rate‑limiting threshold by covering the purpose of throttling, the three core elements of limiting, common algorithms, target dimensions, capacity estimation for single‑service and full‑link scenarios, pressure‑testing techniques, monitoring data, and adaptive configuration strategies.

Capacity PlanningQPSRate Limiting
0 likes · 18 min read
Calculating a 100k QPS Rate‑Limiting Threshold: Methods and Best Practices
Xianyu Technology
Xianyu Technology
Nov 12, 2019 · Cloud Computing

Alibaba's Double 11 2019 Technical Innovations: Live Streaming, AI, Push Platform, and Adaptive Throttling

During Alibaba’s 2019 Double 11, 49 technical teams leveraged cloud‑native systems to hit 10 billion yuan in GMV within 96 seconds and total sales of 268.4 billion yuan, using a sub‑second live‑streaming engine, AI‑driven product recognition, a cloud‑edge push platform that doubled click‑through rates, and an adaptive throttling controller that expanded QPS capacity twentyfold while tripling success rates.

AI Product RecognitionCloud ComputingDouble 11
0 likes · 9 min read
Alibaba's Double 11 2019 Technical Innovations: Live Streaming, AI, Push Platform, and Adaptive Throttling