Evolution of Large-Scale Website Architecture: From Single Server to Distributed Services

After finishing the book Large‑Scale Website Architecture , the author reflects on the concepts of reverse proxy, load balancing, session isolation, and caching, and decides to write a series of blogs to deepen understanding of large‑site technical architecture.

1. Initial Stage Website Architecture

In the early stage a website is often deployed on a single server, as shown in the diagram.

2. Separation of Application and Data Services

As the site grows, a single server can no longer meet demand, so the application layer and the data layer are split into separate servers.

3. Using Caching to Improve Performance

Most modern sites employ caching because roughly 80% of requests target 20% of the data.

4. Application Server Cluster for Concurrency

When a single application server becomes a bottleneck during traffic peaks, a load balancer distributes incoming requests across a cluster of servers.

5. Database Read‑Write Splitting

When user volume grows, the database can become a bottleneck; using master‑slave replication enables read‑write separation to alleviate load.

6. Reverse Proxy and CDN for Faster Responses

Both CDN and reverse proxy rely on caching; CDNs are deployed at ISP data centers, while reverse proxies sit in the site’s own data center, serving cached resources directly to users.

7. Distributed File Systems and Distributed Databases

A single powerful server cannot satisfy the continuous growth of a large site; distributed databases are used as a last resort for extremely large tables, while business‑level sharding is more common.

8. NoSQL and Search Engines

Modern large sites also adopt non‑relational databases such as NoSQL and incorporate search engine technologies for efficient data retrieval.

9. Business Splitting

To handle increasingly complex scenarios, large sites split the overall business into multiple product lines, each deployed as an independent application; communication can be via hyperlinks, message queues, or shared data stores.

10. Distributed Services

Common functionalities such as user management and session handling are extracted into independent services that can be shared across applications.

Source: http://blog.csdn.net/chaofanwei/article/details/26865169

Copyright statement: Content is sourced from the internet and belongs to the original author. We strive to credit the author and source; please notify us of any infringement.