Evolution of Taobao's Technical Architecture and Cloud Migration Best Practices

In its early stage, Taobao quickly launched using the popular LAMP architecture—PHP, Linux, Apache, and MySQL—deployed on about ten application servers with a master‑slave MySQL setup.

By 2004, driven by rapid business growth, Taobao migrated to an Oracle database on IBM mini‑computers with EMC storage, adopting a more expensive but high‑performance solution.

Facing ever‑increasing traffic, the team redesigned the system based on eBay’s architecture, selecting JBoss as the application server, Spring as the IoC container, iBatis for ORM, and building a custom ISearch engine to offload search from Oracle.

In 2006 Taobao built its own CDN to serve static assets (images, descriptions) closer to users, improving response time.

By 2007, with daily transactions exceeding 100 million, Taobao introduced the distributed cache TDBM (predecessor of Tair) to cache hot data such as user profiles and seller ratings, and deployed a self‑developed distributed file system TFS on dozens of x86 servers to replace commercial NAS.

In 2008 the monolithic Oracle architecture was split into more than 20 business centers (product, user, transaction, shop, etc.) using HSF for remote calls and Notify for asynchronous messaging, forming a service‑oriented distributed architecture.

From 2010 onward, Taobao standardized its stack on Alibaba Cloud, leveraging SLB, ECS, RDS, OSS, ONS, CDN and achieving multi‑data‑center disaster recovery and high availability.

The migration to the cloud raised four key technical challenges: availability on PC‑based clusters, consistency compared with Oracle RAC, high I/O performance, and scalable data partitioning.

Best‑practice solutions include stateless applications, extensive caching (browser, reverse‑proxy, page, object, read‑write splitting), service atomization, database sharding, asynchronous processing, minimizing transaction scope, and selective consistency sacrifice, together with automated monitoring and capacity management.

For specific migration scenarios, the article recommends using OSS to replace EMC storage, SLB + multiple ECS instances (or ACE/ONS/OpenSearch) for application services, Alibaba RDS for OLTP workloads, OCS for high‑performance caching, read‑write splitting across multiple RDS instances, horizontal sharding for large tables, and ODPS + OTS + RDS/ADS for OLAP workloads.

Overall, a cloud‑native architecture built on Alibaba Cloud can replace the legacy IOE stack, delivering better performance, scalability, lower cost, and higher availability.