微服务概论(https://microservices.io/)

What are microservices?

Microservices - also known as the microservice architecture - is an architectural style that structures an application as a collection of two or more services that are:

Independently deployable

Loosely coupled

Services are typically organized around business capabilities. Each service is often owned by a single, small team.

微服务只是一种架构风格,它把单体服务拆分成子服务,子服务的特点:

  • 独立开发和部署
  • 松耦合
  • 按照业务能力拆分(DDD)
  • 自治的团队

微服务架构是典型的分治思想。通过对服务的拆分来管理复杂度。避免单体服务的膨胀。

Microservices enable fast flow

微服务通过自治的团队和自动化devops(CI/CD)来实现快速的更新和部署。要实现微服务必须要有对应的组织架构调整(自治的小团队)和基础设施条件。

Design microservices using Assemblage

The key challenge when using microservices is designing a good service architecture.

If you get it wrong you risk creating a distributed monolith, which will slow down software delivery.

Assemblage is an architecture definition process for grouping subdomains/bounded contexts into services.

需要有一套方法来拆分微服务,不然做不到高内聚低耦合,尤其是服务之间以来太强,变成了一个分布式的单体服务

Dark energy and dark matter shape the architecture

Assemblage uses the dark energy and dark matter forces to shape the service architecture.

Dark energy forces encourage decomposition into smaller services. Dark matter forces resist decomposition.

The balance between these forces shapes the service architecture.

基本就是高内聚低耦合的思想。同一领域的业务逻辑应该在一个服务内。subdomain之间的依赖关系最好是单向的。

The pattern language is your guide

The microservices pattern language is your guide when designing an architecture: service collaboration, testing, deployment, common crosscutting concerns and more. It's a collection of patterns that help you make decisions when designing and evolving an architecture.

微服务专业术语使沟通更为便捷

Incrementally refactor a monolith to microservices

The monolithic architecture is not an anti-pattern. It's a good choice for small teams and small projects.

But if you outgrow your monolithic architecture, you need to refactor it to services using the Strangler Fig pattern.

由单体服务过渡到微服务,最重要的是平滑过度。

  • 剥离核心功能
  • 微服务实现核心功能
  • data & api 迁移至微服务
相关推荐
Aethir16 分钟前
去中心化GPU云底层是怎么搭的——去中心化云算力全景·第二章:技术架构
架构·分布式训练·infiniband·nccl·gpu集群·液冷散热
品牌测评2 小时前
Token Plan平台分享|七条算力订阅路径拆解
大数据·人工智能·架构
阿里云云原生2 小时前
2026云栖大会定档!
云原生
人间凡尔赛3 小时前
eBPF + WebAssembly 正在重写服务网格数据平面:2026 云原生架构的“去 Sidecar“革命
后端·云原生·架构
云烟成雨TD5 小时前
Micrometer 系列【33】Spring Boot Micrometer Metrics 自动配置模块解析
spring boot·云原生·micrometer
Dr.kangder5 小时前
嵌入式软件测试(十五)——模型检查技术原理与应用
软件测试·测试工具·架构·嵌入式·静态分析
一次旅行6 小时前
ViT/CLIP/LLaVA/GPT-4V/VideoLLM多模态架构全解:原理仿真+工业落地选型+完整推理代码
人工智能·算法·架构
badhope7 小时前
MCP协议:号称要统一AI工具调用,但大多数人连第一步都走不对
后端·架构
德迅云安全-上官7 小时前
安全加速SCDN与普通CDN深度对比:原理、架构、场景与选型全解析
安全·架构
SLD_Allen7 小时前
Kubernetes + Ray + Volcano:云原生AI训练调度体系
人工智能·云原生·kubernetes