High Concurrency

Prerequisite: Familiarity with the concepts introduced in Executive Summary. Review it first if the terminology in this part is unfamiliar. Answer-first: Handling millions of requests per second (the C10M problem) requires eliminating kernel-space context switching overhead through asynchronous event loops (epoll/kqueue) or kernel-bypass networking (DPDK, io_uring), paired with zero-copy I/O memory buffers, L4 DSR (Direct Server Return) load balancing, and lock-free concurrency structures in Go. flowchart TD Client[Client Traffic Millions req/sec] --> L4[L4 Maglev LB / DPDK DSR] L4 --> L7 Envoy1[L7 Gateway / Envoy Node 1] L4 --> L7 Envoy2[L7 Gateway / Envoy Node 2] subgraph Core Engine [Go High-Concurrency Engine] L7 Envoy1 --> Netpoll[epoll / io_uring Event Loop] Netpoll --> LockFreeQ[Lock-Free Ring Buffer Worker Pool] LockFreeQ --> ZeroCopy[Zero-Copy Memory Allocator sync.Pool] ZeroCopy --> DB[(TiDB / Redis Cluster)] end 1. The Physics of High Concurrency: Beyond C10K to C10M When modern e-commerce platforms like Shopee run Flash Sales or fintech engines like Alipay process Double 11 peak traffic, request rates spike from normal operations (50,000 req/sec) to over 10,000,000 requests per second within milliseconds. ...

Flash Sale Architecture: Rate Limiting & Redis [!NOTE] On sourcing: This article describes flash-sale architecture patterns for C10M-scale events; it is not a disclosure of Shopee’s internal systems, and the figures here are engineering targets rather than published Shopee metrics. Shopee has not publicly documented its flash-sale internals in detail. What is public is its database platform choice — Shopee’s adoption of TiDB is documented in PingCAP’s case studies (How Shopee Chose the Right Database, Shopping on Shopee, the TiDB Way). Treat everything else as a reference pattern to validate against your own workload. ...

← Series hub ← Prev • Next → Answer-first: Alipay’s Double 11 engineering journey evolved over a decade from a centralized monolithic database (2009) to a planet-scale multi-active cloud-native architecture capable of processing over 544,000 TPS at peak. Prerequisite: Executive Summary Overview Answer-first: The Double 11 evolution tracks Alipay journey from monolithic database crashes in 2009 to multi-region active-active unitized architectures. Double 11 (Singles’ Day), initiated in 2009 as a minor promotional event on Taobao Mall, evolved over a decade into the world’s largest online shopping festival. For Alipay’s engineering teams, it served as an annual crucible: a predictable yet extreme spike in transaction volume that forced the continuous redesign of payment infrastructure. This timeline tracks the evolution of Alipay’s technical scaling from a centralized database model to a modern, elastic cloud-native architecture. ...