Product Analysis, Microservice Architecture Design, Real-Time Messaging Evolution, Idempotency, and Service Governance

Author: Architect Lao Lang Link: https://www.jianshu.com/p/19723ab12289

01 Product Analysis and Positioning

02 Microservice Architecture Design

Business‑centric

High cohesion, low coupling

High autonomy

Elastic design

Logging and monitoring

Automation

03 Real‑Time Message Push Technology Evolution

Access‑layer load balancing based on HTTP layer‑7, evolving from HAProxy to Nginx

HAProxy supports TCP and HTTP, offers eight load‑balancing strategies, URL health checks, heartbeat detection, works at layers 4 and 7, but has poor WebSocket support causing message backlog

Nginx supports HTTP, operates at layer‑7, supports WebSocket, port health checks, and powerful regex matching rules

Nginx traffic splitting:

server 172.16.125.76:8066 weight=10;
server 172.16.125.76:8077 down;
server 172.16.0.18:8066 max_fails=3 fail_timeout=30s;
server 172.16.0.18:8077 backup;

HAProxy traffic splitting:

server web01 192.168.137.203:80 check inter 2000 fall 3 weight 10

04 Request Interface Idempotency

Fine‑grained validation with zero intrusion at framework and business layers: filters and interceptors are unsuitable; use annotations with AOP

Duplicate request filtering: AOP around advice checks key existence before proceeding and releases the key after; existing key blocks the request

Concurrent requests: multi‑threaded key queries and creation are unsafe; leverage Redis single‑threaded atomic operations and introduce distributed locks

Atomic key release: only the thread that created the key may release it; handle exceptions in finally block; use Redis transactions with WATCH to monitor the key

Extreme cases: business processing time exceeds key TTL; master‑slave or cluster failover where the slave is not upgraded, causing data loss; introduce Redisson Java solution, periodic key renewal, and distributed in‑memory grid storage

05 Service Governance

Challenges introduced by service‑orientation:

Increasing number of services leads to complex configuration management

Service dependency relationships become intricate

Load balancing among services

Service scaling

Service monitoring

Service degradation

Service authentication

Service rollout and rollback

Service documentation

Comprehensive service governance includes registration & discovery, distributed configuration distribution & hot‑loading, service invocation & load balancing, fault tolerance & degradation, rate limiting, gateway & authentication, intra‑/inter‑network isolation, scheduling, health checks, tracing, business‑level availability checks, continuous delivery, and autonomous operation

-- End --

If you have read this, you probably like this public account; feel free to star (pin) the "Architect Guide" to receive updates promptly.

Reply with "Architect" in the backend of this public account to receive 2 TB of learning materials!

Recommended reading:

Backend Architect Technical Encyclopedia (69 points)

How Should an Architect Design an Authorization System?

How Can I Become an Architect?

Architect Builds an Order System from Scratch