Skip to main content

🧩 Edge Architecture

✅ Overview​

A structural style that performs processing at the "Edge" close to the user, realizing latency reduction and distributed execution.

✅ Problems Addressed​

  • Latency problem by global users.
  • Load concentration on centralized servers.
  • Data explosion due to increase in IoT / Mobile.

✅ Basic Philosophy & Rules​

  • Distributed execution by CDN / Edge Functions / Worker.
  • Authentication, caching, and calculation near the user.
  • Role division with central server (Offloading).

Conceptual Diagram​

✅ Suitable Applications​

  • Web services deploying globally.
  • Local pre-processing of IoT devices.
  • Applications where low latency is important.

❌ Unsuitable Cases​

  • Too much processing placed on the edge (Complication).
  • When there is strict data governance and cannot be distributed.

✅ History​

  • Developed from delivery optimization by CDN.
  • Rapidly spread by Cloudflare Workers / Vercel Edge Functions.

✅ Related Styles​

  • Serverless: Common part with Edge FaaS.
  • Microservices: Affinity as a direction of distributed processing.

✅ Representative Frameworks​

  • Cloudflare Workers / D1 / KV / R2
    Representative Edge Computing environment.

  • Vercel Edge Functions
    Can place SSR / Middleware on the edge.

  • Fastly Compute@Edge
    High-speed Web delivery and edge execution.

  • Akamai EdgeWorkers
    Edge processing platform evolved from CDN.

✅ Design Patterns Supporting This Style​

  • Proxy
    Performs authentication, caching, and filtering at the edge.

  • Facade
    Edge API abstracts the backend.

  • Strategy
    Routing strategy per geography / user type.

  • Observer
    Processing triggered by request events.

✅ Summary​

Edge Architecture is a structural style aiming for:

  • Latency reduction
  • Global distribution
  • Load balancing of central servers

At the same time.

On the other hand, new challenges arise such as:

  • Complexity in deployment
  • Dispersion of logs / traces
  • Data governance and region constraints

When designing cloud frontends or APIs,
it is important to consciously design how much to bring to the edge and from where to leave to the origin.