Skip to content

Latest commit

 

History

History
44 lines (38 loc) · 1.75 KB

yangxin.md

File metadata and controls

44 lines (38 loc) · 1.75 KB

Building Services at Airbnb, Part 4

提高engineering 生产力和质量的挑战

Why schema-oriented infrastructure is critical for service testing

  1. Breaking API Changes
  • API要确保不再被使用才下线。想到了最近我们讨论的方案先@Desprate,几个月后在下线
  • client端先更新deploy，更新API的server端再delpoy
  1. Lack of Usable API Mock Data for Service Consumer
  • 没有组件支持的话，工程师面对不同场景测试需要重复构造大量mock数据
  • 工程师也要体验非常的繁重的setup，这是非常不利于生产的
  1. Lack of Assurance for Mock’s Semantic Correctness
  • 无法保证mock数据语义正确
  • 商业逻辑一直在编，人工维护的mock可能会过时
  1. Lack of Validation for Service Owner
  • 框架要支持Validation
  1. Lack of Real Time Metrics for API Test Quality
  • test coverage and test data quality

How service owners and consumers benefit

提供了一些框架

  1. Static API Schema Validation

  • Backwards Compatibility Check focuses on detecting potential API breaking changes.

  • Schema Linter aims at enforcing our IDL best practices.

  • 禁止不向后兼容的schema修改

  • 实现 Abstract Syntax Tree

  • 接入CICD

  1. Schema-based API Data Factory

  • Mocking dependent service API in unit tests and integration tests.

  • Validating API endpoint correctness using the materialized API request/response.

  • mock数据与语言无关，通用

  • CI检测数据是否与最新API兼容

  1. API Mocking Framework
  • 模拟调用
  • latency
  1. API Integration Testing Framework benefit
  • Producer service is validated without writing boilerplate code
  • Mocking data is semantically validated
  • Provide realtime API test coverage instrumentation
  • Pluggability and Lightweight