Meta is secretly testing an AI shopping assistant that leverages billions of users' social profiles, shifting recommendation logic from reactive behavior data to proactive identity‑driven suggestions, while raising significant privacy and ecosystem implications for e‑commerce.
This article details the design and implementation of REDtao, a self‑built graph storage system for Xiaohongshu that replaces MySQL with a three‑layer architecture, distributed cache, cross‑cloud multi‑active support, and delivers trillion‑edge scale, 150 M QPS, 90% cache hit rate, and significant cost reductions.
Xiaohongshu built the REDtao graph storage system to handle a trillion‑scale social graph, replacing MySQL with a three‑layer architecture, custom graph APIs, high‑availability caches, cross‑cloud multi‑active deployment, and cloud‑native operators, achieving over 90% cache hit rate and dramatic cost savings.
REDtao is a high‑performance graph storage system built for Xiaohongshu that extends Facebook’s Tao architecture with a three‑layer hash structure, decoupled caching, leader‑follower distribution and cross‑cloud availability, delivering over 90% cache hits, 70% MySQL CPU reduction, 150 M QPS on a 16‑core server and seamless migration despite a 250% surge in daily‑active‑user traffic.
