Zuul2 超時設置 重試 踩坑

背景：

承接上一篇博客，針對的ZUUL2的超時 重試機制，踩坑了一下

配置如下

logging:
  level:
    root: info

demo-service:
  ribbon:
#      ReadTimeout: 7000
#      ConnectTimeout: 5800
#      MaxAutoRetries: 3
#      MaxAutoRetriesNextServer: 0
#      retryableStatusCodes: 500,502,503,404
      listOfServers: localhost:8002
#      OkToRetryOnAllOperations: true


zuul:
  retryable: true
  eureka:
    client:
      enabled: false
  routes:
    demo-service:
      path: /service/**
      serviceId: demo-service
#  host:
#    socket-timeout-millis: 5000
#    connect-timeout-millis: 4000
  sensitive-headers: Access-Control-Allow-Origin,Access-Control-Allow-Methods

ribbon:
  ReadTimeout: 5000
  ConnectTimeout: 1000
  MaxAutoRetries: 5
  MaxAutoRetriesNextServer: 0
#  retryableStatusCodes: 500,502,503

坑點:

• 通過源碼調試，在 HttpClientRibbonConfiguration類的retryableRibbonLoadBalancingHttpClient方法，可以看到默認的RetryableRibbonLoadBalancingHttpClient實例，主要屬性的大小寫，不然不會生效
• ConnectTimeout 實際生效數值是配置的2倍，所以你配置1秒，實際是2秒，切記!!!
• 記得引入spring-retry組件，不然重試不會生效