【云原生 | Envoy 系列】--Envoy 动态配置及配置源-基于文件系统订阅

企业开发 2022-10-25 01:03:03 阅读次数: 0

1. Envoy 动态配置及配置源

1.1 Envoy 动态配置

  • xDS API为Envoy提供了资源的动态配置机制,它也被称为Data Plane API(遵循UPDA规范)
  • Envoy支持三种类型的配置信息的动态发现机制,相关的发现服务及相应的API联合起来称为xDS API
    • 基于文件发现: 指定要监视的文件系统路径
    • 通过查询一到多个管理服务区(Manager Server)发现: 通过DiscoveryRequest协议报文发送请求,并要求服务方以DiscoveryResponse协议报文进行响应.
      • gRPC服务: 启动gRPC流(长连接)
      • REST服务: 轮询REST-JSON URL(轮询间隔一定时间访问一次REST)
  • v3 xDS支持如下几种资源类型
    • envoy.config.listener.v3.Listener
    • envoy.config.route.v3.RouteConfiguration
    • envoy.config.route.v3.ScopedRouteConfiguration
    • envoy.config.route.v3.VirtualHost
    • envoy.config.cluster.v3.Cluster
    • envoy.config.endpoint.v3.ClusterLoadAssignment
    • envoy.extensions.transport_sockets.tls.v3.Secret
    • envoy.service.runtime.v3.Runtime

1.2 xDS API概述

Envoy对xDS API的管理由后端服务器实现,包括LDS,CDS,RDS,SRDS(Scpoed Route),VHDS,EDS,SDS,RTDS等

  • 所有这些API都提供了最终的一致性.并彼此之间不存在相互影响
  • 部分高级操作需要进行排序以防止流量被丢弃,因此,基于一个管理服务器提供多类API时还需要使用聚合发现服务(ADS)API
    ADS API允许所有其他API通过来自单个管理服务器的单个gRPC双向流进行编组,从而允许对操作进行确定性排序
  • xDS的各API还支持增量传输机制,包括ADS

1.3 Bootstrap node 配置

一个Managerment Server实例可能需要同时响应多个不同的Envoy实例的资源发现请求

  • Managerment Server上的配置需要为适配到不同的Envoy实例
  • Envoy实例请求发现配置时,需要在请求报文中上报自身的信息
    • id,cluster,metadata,locality等
    • 这些配置信息定义在Bootstrap配置文件中
node:
  id:				# 必选
  cluster:			# 必选
  metadata: {
    
    ...}		# 非必选
  locality:			    # 非必选
    region:
    zone:
    sub_zone:
  user_agent_name:			# 向上游模拟身份
  user_agent_version:
  user_agent_build_version:
    version:
    metadata: {
    
    ...}
  extensions: []
  client_features: []
  listening_addresses: []

1.4 API基本工作流程

对于经典HTTP路由方案,xDS API的Managerment Server需要为其客户端(Envoy实例)配置的核心资源类型为Listener,RouteConfiguration,Cluster和ClusterLoadAssignment(endpoint)四个(还有Virtual Host)

  • 每个Listener资源可以指向一个RouteConfiguration资源,该资源可以指向一个或多个Cluster资源,并且每一个Cluster资源可以指向一个ClusterLoadAssignment资源

Envoy实例启动时请求加载所有Listener和Cluster资源,而后再获取这些Listener和Cluster所依赖的RouteConfiguration和ClusterLoadAssignment配置

  • Listener资源和Cluster资源分别代表着客户端上的根(root)配置,并且可并行加载

类型gRPC一类的非代理式客户端可以仅在启动是请求加载其感兴趣的Listener资源,而后再加载这些特定的Listener相关的RouteConfiguration资源,最后这些RouteConfiguration资源指向的Cluster资源,以及由这些cluster资源依赖的ClusterLoadAssignment资源

  • Listener资源是客户端整个配置的根

1.5 Envoy 资源的配置源(ConfigSource)

配置源(ConfigSource)用于指定资源配置数据的来源,用于为Listener,Cluster,Route,Endpoint,Secret和Virtual Host等资源提供配置数据

目前,Envoy支持的资源配置源只能是path,api_config_souorce或ads其中之一

api_config_source或ads的数据来源于xDS API Server,即Manager Server

2. 基于文件系统的订阅

为Envoy提供动态配置的最简单方法是将其放置在ConfigSource中显示指定的文件路径中

  • Envoy将使用inotify来监视文件的更改,并在更新时解析文件中的DiscoveryResponse报文
  • 二进制protobufs,Json,Yaml和proto文本都是DiscoveryResponse所支持的数据格式
  • 除了统计计数日志外,没有任何机制可用于文件系统订阅ACK/NACK更新
  • 若配置更新被拒绝,xDS API的最后一个有效配置将继续使用

2.1 基于文件系统EDS动态实现

2.1.1 Docker-compose.yaml

五个Service:

  • envoy: Front Proxy,地址为172.31.11.2
  • webserver01: 第一个后端服务
  • webserver01-sidecat: 第一个后端服务的sidecar Proxy,地址为172.31.11.11
  • webserver02: 第二个后端服务
  • webserver02-sidecar: 第二个后端服务的sidecar Proxy,地址为172.31.11.12
version: '3.3'

services:
  envoy:
    image: envoyproxy/envoy-alpine:v1.21.5
    environment:
      - ENVOY_UID=0
      - ENVOY_GID=0
    volumes:
    - ./front-envoy.yaml:/etc/envoy/envoy.yaml  ## 将本地的front-envoy.yaml映射给/etc/envoy/envoy.yaml
    - ./eds.conf.d/:/etc/envoy/eds.conf.d/  # 将本地eds.conf.d目录下的文件映射给/etc/envoy/eds.conf.d/目录
    networks:
      envoymesh:
        ipv4_address: 172.31.11.2
        aliases:
        - front-proxy
    depends_on:
    - webserver01-sidecar
    - webserver02-sidecar

  webserver01-sidecar:
    image: envoyproxy/envoy-alpine:v1.21.5
    volumes:
    - ./envoy-sidecar-proxy.yaml:/etc/envoy/envoy.yaml
    hostname: webserver01
    networks:
      envoymesh:
        ipv4_address: 172.31.11.11
        aliases:
        - webserver01-sidecar

  webserver01:
    image: ikubernetes/demoapp:v1.0
    environment:
      - PORT=8080
      - HOST=127.0.0.1
    network_mode: "service:webserver01-sidecar"
    depends_on:
    - webserver01-sidecar

  webserver02-sidecar:
    image: envoyproxy/envoy-alpine:v1.21.5
    volumes:
    - ./envoy-sidecar-proxy.yaml:/etc/envoy/envoy.yaml
    hostname: webserver02
    networks:
      envoymesh:
        ipv4_address: 172.31.11.12
        aliases:
        - webserver02-sidecar

  webserver02:
    image: ikubernetes/demoapp:v1.0
    environment:
      - PORT=8080
      - HOST=127.0.0.1
    network_mode: "service:webserver02-sidecar"
    depends_on:
    - webserver02-sidecar

networks:
  envoymesh:
    driver: bridge
    ipam:
      config:
        - subnet: 172.31.11.0/24

2.1.2 front envoy.yaml

node:
  id: envoy_front_proxy		# 说明自己是谁
  cluster: pana_Cluster		# 说明自己是哪个集群

admin:				# 打开admin接口,方便后面通过curl查询集群信息
  profile_path: /tmp/envoy.prof
  access_log_path: /tmp/admin_access.log
  address:
    socket_address:
       address: 0.0.0.0
       port_value: 9901

static_resources:
  listeners:
  - name: listener_0
    address:
      socket_address: {
    
     address: 0.0.0.0, port_value: 80 }
    filter_chains:
    - filters:
      - name: envoy.filters.network.http_connection_manager
        typed_config:
          "@type": type.googleapis.com/envoy.extensions.filters.network.http_connection_manager.v3.HttpConnectionManager
          stat_prefix: ingress_http
          codec_type: AUTO
          route_config:
            name: local_route
            virtual_hosts:
            - name: web_service_01
              domains: ["*"]
              routes:
              - match: {
    
     prefix: "/" }
                route: {
    
     cluster: webcluster }
          http_filters:
          - name: envoy.filters.http.router

  clusters:
  - name: webcluster
    connect_timeout: 0.25s
    type: EDS					# 动态eds配置方式
    lb_policy: ROUND_ROBIN
    eds_cluster_config:
      service_name: webcluster
      eds_config:
        path: '/etc/envoy/eds.conf.d/eds.yaml'   # 配置文件为/etc/envoy/eds.conf.d/eds.yaml

2.1.3 Sidecar Envoy

admin:
  profile_path: /tmp/envoy.prof
  access_log_path: /tmp/admin_access.log
  address:
    socket_address:
       address: 0.0.0.0
       port_value: 9901

static_resources:
  listeners:
  - name: listener_0
    address:
      socket_address: {
    
     address: 0.0.0.0, port_value: 80 }
    filter_chains:
    - filters:
      - name: envoy.filters.network.http_connection_manager
        typed_config:
          "@type": type.googleapis.com/envoy.extensions.filters.network.http_connection_manager.v3.HttpConnectionManager
          stat_prefix: ingress_http
          codec_type: AUTO
          route_config:
            name: local_route
            virtual_hosts:
            - name: local_service 
              domains: ["*"]
              routes:
              - match: {
    
     prefix: "/" }
                route: {
    
     cluster: local_cluster }
          http_filters:
          - name: envoy.filters.http.router

  clusters:
  - name: local_cluster
    connect_timeout: 0.25s
    type: STATIC
    lb_policy: ROUND_ROBIN
    load_assignment:
      cluster_name: local_cluster
      endpoints:
      - lb_endpoints:
        - endpoint:
            address:
              socket_address: {
    
     address: 127.0.0.1, port_value: 8080 }

2.1.4 eds.yaml

默认配置中只有172.31.11.11的配置

# cat eds.yaml
resources:
- "@type": type.googleapis.com/envoy.config.endpoint.v3.ClusterLoadAssignment
  cluster_name: webcluster
  endpoints:
  - lb_endpoints:
    - endpoint:
        address:
          socket_address:
            address: 172.31.11.11
            port_value: 80

2.1.5 运行测试

  1. 默认情况下,cluster中只包含了172.31.11.11:80
# docker-compose up
## 此时可以看到默认情况下,cluster中只包含了172.31.11.11:80
# curl 172.31.11.2:9901/clusters
webcluster::observability_name::webcluster
webcluster::default_priority::max_connections::1024
webcluster::default_priority::max_pending_requests::1024
webcluster::default_priority::max_requests::1024
webcluster::default_priority::max_retries::3
webcluster::high_priority::max_connections::1024
webcluster::high_priority::max_pending_requests::1024
webcluster::high_priority::max_requests::1024
webcluster::high_priority::max_retries::3
webcluster::added_via_api::false
webcluster::172.31.11.11:80::cx_active::0
webcluster::172.31.11.11:80::cx_connect_fail::0
webcluster::172.31.11.11:80::cx_total::0
webcluster::172.31.11.11:80::rq_active::0
webcluster::172.31.11.11:80::rq_error::0
webcluster::172.31.11.11:80::rq_success::0
webcluster::172.31.11.11:80::rq_timeout::0
webcluster::172.31.11.11:80::rq_total::0
webcluster::172.31.11.11:80::hostname::
webcluster::172.31.11.11:80::health_flags::healthy
webcluster::172.31.11.11:80::weight::1
webcluster::172.31.11.11:80::region::
webcluster::172.31.11.11:80::zone::
webcluster::172.31.11.11:80::sub_zone::
webcluster::172.31.11.11:80::canary::false
webcluster::172.31.11.11:80::priority::0
webcluster::172.31.11.11:80::success_rate::-1.0
webcluster::172.31.11.11:80::local_origin_success_rate::-1.0
## 进入容器
# docker exec -it c73086aa69f7 sh
/ # cd /etc/envoy/eds.conf.d
/etc/envoy/eds.conf.d # cat eds.yaml
resources:
- "@type": type.googleapis.com/envoy.config.endpoint.v3.ClusterLoadAssignment
  cluster_name: webcluster
  endpoints:
  - lb_endpoints:
    - endpoint:
        address:
          socket_address:
            address: 172.31.11.11
            port_value: 80
## 给eds.yaml追加配置后
/etc/envoy/eds.conf.d # cat eds.yaml
version_info: '2'
resources:
- "@type": type.googleapis.com/envoy.config.endpoint.v3.ClusterLoadAssignment
  cluster_name: webcluster
  endpoints:
  - lb_endpoints:
    - endpoint:
        address:
          socket_address:
            address: 172.31.11.11
            port_value: 80
    - endpoint:
        address:
          socket_address:
            address: 172.31.11.12
            port_value: 80
## 强行改名后触发配置生效
/etc/envoy/eds.conf.d # mv eds.yaml tmp && mv tmp eds.yaml
## 此时在docker容器之外可以看到,集群下除了172.31.11.11还有172.31.11.12
# curl 172.31.11.2:9901/clusters
webcluster::observability_name::webcluster
webcluster::default_priority::max_connections::1024
webcluster::default_priority::max_pending_requests::1024
webcluster::default_priority::max_requests::1024
webcluster::default_priority::max_retries::3
webcluster::high_priority::max_connections::1024
webcluster::high_priority::max_pending_requests::1024
webcluster::high_priority::max_requests::1024
webcluster::high_priority::max_retries::3
webcluster::added_via_api::false
webcluster::172.31.11.11:80::cx_active::0
webcluster::172.31.11.11:80::cx_connect_fail::0
webcluster::172.31.11.11:80::cx_total::0
webcluster::172.31.11.11:80::rq_active::0
webcluster::172.31.11.11:80::rq_error::0
webcluster::172.31.11.11:80::rq_success::0
webcluster::172.31.11.11:80::rq_timeout::0
webcluster::172.31.11.11:80::rq_total::0
webcluster::172.31.11.11:80::hostname::
webcluster::172.31.11.11:80::health_flags::healthy
webcluster::172.31.11.11:80::weight::1
webcluster::172.31.11.11:80::region::
webcluster::172.31.11.11:80::zone::
webcluster::172.31.11.11:80::sub_zone::
webcluster::172.31.11.11:80::canary::false
webcluster::172.31.11.11:80::priority::0
webcluster::172.31.11.11:80::success_rate::-1.0
webcluster::172.31.11.11:80::local_origin_success_rate::-1.0
webcluster::172.31.11.12:80::cx_active::0
webcluster::172.31.11.12:80::cx_connect_fail::0
webcluster::172.31.11.12:80::cx_total::0
webcluster::172.31.11.12:80::rq_active::0
webcluster::172.31.11.12:80::rq_error::0
webcluster::172.31.11.12:80::rq_success::0
webcluster::172.31.11.12:80::rq_timeout::0
webcluster::172.31.11.12:80::rq_total::0
webcluster::172.31.11.12:80::hostname::
webcluster::172.31.11.12:80::health_flags::healthy
webcluster::172.31.11.12:80::weight::1
webcluster::172.31.11.12:80::region::
webcluster::172.31.11.12:80::zone::
webcluster::172.31.11.12:80::sub_zone::
webcluster::172.31.11.12:80::canary::false
webcluster::172.31.11.12:80::priority::0
webcluster::172.31.11.12:80::success_rate::-1.0
webcluster::172.31.11.12:80::local_origin_success_rate::-1.0
## 尝试请求listener,会分别向2个ep调度请求
root@k8s-node-1:~# curl 172.31.11.2
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver01, ServerIP: 172.31.11.11!
root@k8s-node-1:~# curl 172.31.11.2
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver01, ServerIP: 172.31.11.11!
root@k8s-node-1:~# curl 172.31.11.2
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver02, ServerIP: 172.31.11.12!
root@k8s-node-1:~# curl 172.31.11.2
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver01, ServerIP: 172.31.11.11!
## 当webserver1出现故障时,对文件进行修改移除故障的endpoint
/etc/envoy/eds.conf.d # cat eds.yaml
version_info: '2'
resources:
- "@type": type.googleapis.com/envoy.config.endpoint.v3.ClusterLoadAssignment
  cluster_name: webcluster
  endpoints:
  - lb_endpoints:
    - endpoint:
        address:
          socket_address:
            address: 172.31.11.12
            port_value: 80
## 刷新服务
/etc/envoy/eds.conf.d # mv eds.yaml bak && mv bak eds.yaml
## 此时查看clusters的情况可以看到172.31.11.11已经被移除
# curl 172.31.11.2:9901/clusters
webcluster::observability_name::webcluster
webcluster::default_priority::max_connections::1024
webcluster::default_priority::max_pending_requests::1024
webcluster::default_priority::max_requests::1024
webcluster::default_priority::max_retries::3
webcluster::high_priority::max_connections::1024
webcluster::high_priority::max_pending_requests::1024
webcluster::high_priority::max_requests::1024
webcluster::high_priority::max_retries::3
webcluster::added_via_api::false
webcluster::172.31.11.12:80::cx_active::0
webcluster::172.31.11.12:80::cx_connect_fail::0
webcluster::172.31.11.12:80::cx_total::0
webcluster::172.31.11.12:80::rq_active::0
webcluster::172.31.11.12:80::rq_error::0
webcluster::172.31.11.12:80::rq_success::0
webcluster::172.31.11.12:80::rq_timeout::0
webcluster::172.31.11.12:80::rq_total::0
webcluster::172.31.11.12:80::hostname::
webcluster::172.31.11.12:80::health_flags::healthy
webcluster::172.31.11.12:80::weight::1
webcluster::172.31.11.12:80::region::
webcluster::172.31.11.12:80::zone::
webcluster::172.31.11.12:80::sub_zone::
webcluster::172.31.11.12:80::canary::false
webcluster::172.31.11.12:80::priority::0
webcluster::172.31.11.12:80::success_rate::-1.0
webcluster::172.31.11.12:80::local_origin_success_rate::-1.0
## 访问集群只会被调度到172.31.11.12上了
# curl 172.31.11.2
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver02, ServerIP: 172.31.11.12!
# curl 172.31.11.2
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver02, ServerIP: 172.31.11.12!
# curl 172.31.11.2
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver02, ServerIP: 172.31.11.12!
# curl 172.31.11.2
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver02, ServerIP: 172.31.11.12!

通过修改eds.yaml文件,实现后端endpoint动态增删

请添加图片描述

至此,通过文件的动态修改实现了对后端EP的增删.这就是EDS

2.2 基于文件系统LDS,CDS动态实现

2.2.1 Docker-compose

五个Service:

  • envoy: Front Proxy,地址为172.31.12.2
  • webserver01: 第一个后端服务
  • webserver01-sidecat: 第一个后端服务的sidecar Proxy,地址为172.31.12.11
  • webserver02: 第二个后端服务
  • webserver02-sidecar: 第二个后端服务的sidecar Proxy,地址为172.31.12.12
version: '3.3'

services:
  envoy:
    image: envoyproxy/envoy-alpine:v1.21.5
    environment:
      - ENVOY_UID=0
      - ENVOY_GID=0
    volumes:
    - ./front-envoy.yaml:/etc/envoy/envoy.yaml
    - ./eds.conf.d/:/etc/envoy/eds.conf.d/
    networks:
      envoymesh:
        ipv4_address: 172.31.11.2
        aliases:
        - front-proxy
    depends_on:
    - webserver01-sidecar
    - webserver02-sidecar

  webserver01-sidecar:
    image: envoyproxy/envoy-alpine:v1.21.5
    volumes:
    - ./envoy-sidecar-proxy.yaml:/etc/envoy/envoy.yaml
    hostname: webserver01
    networks:
      envoymesh:
        ipv4_address: 172.31.11.11
        aliases:
        - webserver01-sidecar
        - webserver

  webserver01:
    image: ikubernetes/demoapp:v1.0
    environment:
      - PORT=8080
      - HOST=127.0.0.1
    network_mode: "service:webserver01-sidecar"
    depends_on:
    - webserver01-sidecar

  webserver02-sidecar:
    image: envoyproxy/envoy-alpine:v1.21.5
    volumes:
    - ./envoy-sidecar-proxy.yaml:/etc/envoy/envoy.yaml
    hostname: webserver02
    networks:
      envoymesh:
        ipv4_address: 172.31.11.12
        aliases:
        - webserver02-sidecar
        - webserver

  webserver02:
    image: ikubernetes/demoapp:v1.0
    environment:
      - PORT=8080
      - HOST=127.0.0.1
    network_mode: "service:webserver02-sidecar"
    depends_on:
    - webserver02-sidecar

networks:
  envoymesh:
    driver: bridge
    ipam:
      config:
        - subnet: 172.31.11.0/24

2.2.2 front envoy.yaml

这里就相对之前的配置简单很多,除了node的id标记和用于查询admin之外,其他配置只有5行.

通过发现L(listener)发现下面的VirtualHost,Route配置

通过Route调用Client

通过发现Client,发现下面的Endpoint配置

所以这里只需要指定L和C即可

node:
  id: envoy_front_proxy
  cluster: MageEdu_Cluster

admin:
  profile_path: /tmp/envoy.prof
  access_log_path: /tmp/admin_access.log
  address:
    socket_address:
       address: 0.0.0.0
       port_value: 9901

dynamic_resources:
  lds_config:
    path: /etc/envoy/conf.d/lds.yaml
  cds_config:
    path: /etc/envoy/conf.d/cds.yaml

2.2.3 lds.yaml

这部分的设定和之前静态的配置内容几乎相同,将后端cluster转发到webcluster

resources:
- "@type": type.googleapis.com/envoy.config.listener.v3.Listener
  name: listener_http
  address:
    socket_address: {
    
     address: 0.0.0.0, port_value: 80 }
  filter_chains:
  - filters:
      name: envoy.http_connection_manager
      typed_config:
        "@type": type.googleapis.com/envoy.extensions.filters.network.http_connection_manager.v3.HttpConnectionManager
        stat_prefix: ingress_http
        route_config:
          name: local_route
          virtual_hosts:
          - name: local_service
            domains: ["*"]
            routes:
            - match:
                prefix: "/"
              route:
                cluster: webcluster
        http_filters:
        - name: envoy.filters.http.router

2.2.4 cds.yaml

这里type设置为了静态.也可以将type设置成EDS,在使用eds_config去做endpoint的动态发现

resources:
- "@type": type.googleapis.com/envoy.config.cluster.v3.Cluster
  name: webcluster
  connect_timeout: 1s
  type: STRICT_DNS
  load_assignment:
    cluster_name: webcluster
    endpoints:
    - lb_endpoints:
      - endpoint:
          address:
            socket_address:
              address: webserver01
              port_value: 80
      - endpoint:
          address:
            socket_address:
              address: webserver02
              port_value: 80

2.2.5 运行测试

# docker-compose up
## 可以看到在12.11和12.12之间轮询
# curl 172.31.12.2
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver01, ServerIP: 172.31.12.11!
# curl 172.31.12.2
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver02, ServerIP: 172.31.12.12!
# curl 172.31.12.2
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver01, ServerIP: 172.31.12.11!
# curl 172.31.12.2
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver02, ServerIP: 172.31.12.12!
# curl 172.31.12.2
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver01, ServerIP: 172.31.12.11!
## cluster中也可以看到2个endpoint
# curl 172.31.12.2:9901/clusters
webcluster::observability_name::webcluster
webcluster::default_priority::max_connections::1024
webcluster::default_priority::max_pending_requests::1024
webcluster::default_priority::max_requests::1024
webcluster::default_priority::max_retries::3
webcluster::high_priority::max_connections::1024
webcluster::high_priority::max_pending_requests::1024
webcluster::high_priority::max_requests::1024
webcluster::high_priority::max_retries::3
webcluster::added_via_api::true
webcluster::172.31.12.11:80::cx_active::2
webcluster::172.31.12.11:80::cx_connect_fail::0
webcluster::172.31.12.11:80::cx_total::2
webcluster::172.31.12.11:80::rq_active::0
webcluster::172.31.12.11:80::rq_error::0
webcluster::172.31.12.11:80::rq_success::4
webcluster::172.31.12.11:80::rq_timeout::0
webcluster::172.31.12.11:80::rq_total::4
webcluster::172.31.12.11:80::hostname::webserver01
webcluster::172.31.12.11:80::health_flags::healthy
webcluster::172.31.12.11:80::weight::1
webcluster::172.31.12.11:80::region::
webcluster::172.31.12.11:80::zone::
webcluster::172.31.12.11:80::sub_zone::
webcluster::172.31.12.11:80::canary::false
webcluster::172.31.12.11:80::priority::0
webcluster::172.31.12.11:80::success_rate::-1.0
webcluster::172.31.12.11:80::local_origin_success_rate::-1.0
webcluster::172.31.12.12:80::cx_active::1
webcluster::172.31.12.12:80::cx_connect_fail::0
webcluster::172.31.12.12:80::cx_total::1
webcluster::172.31.12.12:80::rq_active::0
webcluster::172.31.12.12:80::rq_error::0
webcluster::172.31.12.12:80::rq_success::2
webcluster::172.31.12.12:80::rq_timeout::0
webcluster::172.31.12.12:80::rq_total::2
webcluster::172.31.12.12:80::hostname::webserver02
webcluster::172.31.12.12:80::health_flags::healthy
webcluster::172.31.12.12:80::weight::1
webcluster::172.31.12.12:80::region::
webcluster::172.31.12.12:80::zone::
webcluster::172.31.12.12:80::sub_zone::
webcluster::172.31.12.12:80::canary::false
webcluster::172.31.12.12:80::priority::0
webcluster::172.31.12.12:80::success_rate::-1.0
webcluster::172.31.12.12:80::local_origin_success_rate::-1.0
# curl 172.31.12.2:9901/listeners
listener_http::0.0.0.0:80

2.2.5.1 测试修改lds

将listener的监听由80改为8081,并使用mv触发变更生效

resources:
- "@type": type.googleapis.com/envoy.config.listener.v3.Listener
  name: listener_http
  address:
    socket_address: {
    
     address: 0.0.0.0, port_value: 8081 }
  filter_chains:
  - filters:
      name: envoy.http_connection_manager
      typed_config:
        "@type": type.googleapis.com/envoy.extensions.filters.network.http_connection_manager.v3.HttpConnectionManager
        stat_prefix: ingress_http
        route_config:
          name: local_route
          virtual_hosts:
          - name: local_service
            domains: ["*"]
            routes:
            - match:
                prefix: "/"
              route:
                cluster: webcluster
        http_filters:
        - name: envoy.filters.http.router
/etc/envoy/conf.d # mv lds.yaml tmp && mv tmp lds.yaml

可以看到监听 已经变为了8081,尝试访问8081

root@k8s-node-1:~# curl 172.31.12.2:9901/listeners
listener_http::0.0.0.0:8081
root@k8s-node-1:~# curl 172.31.12.2:8081
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver01, ServerIP: 172.31.12.11!
root@k8s-node-1:~# curl 172.31.12.2:8081
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver02, ServerIP: 172.31.12.12!
root@k8s-node-1:~# curl 172.31.12.2:8081
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver02, ServerIP: 172.31.12.12!
root@k8s-node-1:~# curl 172.31.12.2:8081
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver01, ServerIP: 172.31.12.11!

2.2.5.2 测试修改cds

在cds中删除12服务部分

resources:
- "@type": type.googleapis.com/envoy.config.cluster.v3.Cluster
  name: webcluster
  connect_timeout: 1s
  type: STRICT_DNS
  load_assignment:
    cluster_name: webcluster
    endpoints:
    - lb_endpoints:
      - endpoint:
          address:
            socket_address:
              address: webserver01
              port_value: 80
## 触发配置生效
/etc/envoy/conf.d # mv cds.yaml tmp&& mv tmp cds.yaml

此时集群中endpoint只有11了

root@k8s-node-1:~# curl 172.31.12.2:9901/clusters
webcluster::observability_name::webcluster
webcluster::default_priority::max_connections::1024
webcluster::default_priority::max_pending_requests::1024
webcluster::default_priority::max_requests::1024
webcluster::default_priority::max_retries::3
webcluster::high_priority::max_connections::1024
webcluster::high_priority::max_pending_requests::1024
webcluster::high_priority::max_requests::1024
webcluster::high_priority::max_retries::3
webcluster::added_via_api::true
webcluster::172.31.12.11:80::cx_active::0
webcluster::172.31.12.11:80::cx_connect_fail::0
webcluster::172.31.12.11:80::cx_total::0
webcluster::172.31.12.11:80::rq_active::0
webcluster::172.31.12.11:80::rq_error::0
webcluster::172.31.12.11:80::rq_success::0
webcluster::172.31.12.11:80::rq_timeout::0
webcluster::172.31.12.11:80::rq_total::0
webcluster::172.31.12.11:80::hostname::webserver01
webcluster::172.31.12.11:80::health_flags::healthy
webcluster::172.31.12.11:80::weight::1
webcluster::172.31.12.11:80::region::
webcluster::172.31.12.11:80::zone::
webcluster::172.31.12.11:80::sub_zone::
webcluster::172.31.12.11:80::canary::false
webcluster::172.31.12.11:80::priority::0
webcluster::172.31.12.11:80::success_rate::-1.0
webcluster::172.31.12.11:80::local_origin_success_rate::-1.0
root@k8s-node-1:~# curl 172.31.12.2:8081
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver01, ServerIP: 172.31.12.11!
root@k8s-node-1:~# curl 172.31.12.2:8081
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver01, ServerIP: 172.31.12.11!
root@k8s-node-1:~# curl 172.31.12.2:8081
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver01, ServerIP: 172.31.12.11!
root@k8s-node-1:~# curl 172.31.12.2:8081
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver01, ServerIP: 172.31.12.11!

2.2.5.3 再次修改cds

由于cds配置的是STRICT_DNS,STRICT_DNS是依次解析出所有域名中的ip

又由于11和12都有别名webserver

        ipv4_address: 172.31.11.11
        aliases:
        - webserver01-sidecar
        - webserver
        ....
        ipv4_address: 172.31.11.12
        aliases:
        - webserver02-sidecar
        - webserver

将配置文件的address改为了webserver,并重新触发配置生效

resources:
- "@type": type.googleapis.com/envoy.config.cluster.v3.Cluster
  name: webcluster
  connect_timeout: 1s
  type: STRICT_DNS
  load_assignment:
    cluster_name: webcluster
    endpoints:
    - lb_endpoints:
      - endpoint:
          address:
            socket_address:
              address: webserver
              port_value: 80
/etc/envoy/conf.d # mv cds.yaml tmp && mv tmp cds.yaml

此时cluster有了11和12的ep

root@k8s-node-1:~# curl 172.31.12.2:9901/clusters
webcluster::observability_name::webcluster
webcluster::default_priority::max_connections::1024
webcluster::default_priority::max_pending_requests::1024
webcluster::default_priority::max_requests::1024
webcluster::default_priority::max_retries::3
webcluster::high_priority::max_connections::1024
webcluster::high_priority::max_pending_requests::1024
webcluster::high_priority::max_requests::1024
webcluster::high_priority::max_retries::3
webcluster::added_via_api::true
webcluster::172.31.12.11:80::cx_active::0
webcluster::172.31.12.11:80::cx_connect_fail::0
webcluster::172.31.12.11:80::cx_total::0
webcluster::172.31.12.11:80::rq_active::0
webcluster::172.31.12.11:80::rq_error::0
webcluster::172.31.12.11:80::rq_success::0
webcluster::172.31.12.11:80::rq_timeout::0
webcluster::172.31.12.11:80::rq_total::0
webcluster::172.31.12.11:80::hostname::webserver
webcluster::172.31.12.11:80::health_flags::healthy
webcluster::172.31.12.11:80::weight::1
webcluster::172.31.12.11:80::region::
webcluster::172.31.12.11:80::zone::
webcluster::172.31.12.11:80::sub_zone::
webcluster::172.31.12.11:80::canary::false
webcluster::172.31.12.11:80::priority::0
webcluster::172.31.12.11:80::success_rate::-1.0
webcluster::172.31.12.11:80::local_origin_success_rate::-1.0
webcluster::172.31.12.12:80::cx_active::0
webcluster::172.31.12.12:80::cx_connect_fail::0
webcluster::172.31.12.12:80::cx_total::0
webcluster::172.31.12.12:80::rq_active::0
webcluster::172.31.12.12:80::rq_error::0
webcluster::172.31.12.12:80::rq_success::0
webcluster::172.31.12.12:80::rq_timeout::0
webcluster::172.31.12.12:80::rq_total::0
webcluster::172.31.12.12:80::hostname::webserver
webcluster::172.31.12.12:80::health_flags::healthy
webcluster::172.31.12.12:80::weight::1
webcluster::172.31.12.12:80::region::
webcluster::172.31.12.12:80::zone::
webcluster::172.31.12.12:80::sub_zone::
webcluster::172.31.12.12:80::canary::false
webcluster::172.31.12.12:80::priority::0
webcluster::172.31.12.12:80::success_rate::-1.0
webcluster::172.31.12.12:80::local_origin_success_rate::-1.0

再次访问listener,可以在2个endpoint中进行轮询

root@k8s-node-1:~# curl 172.31.12.2:8081
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver01, ServerIP: 172.31.12.11!
root@k8s-node-1:~# curl 172.31.12.2:8081
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver01, ServerIP: 172.31.12.11!
root@k8s-node-1:~# curl 172.31.12.2:8081
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver02, ServerIP: 172.31.12.12!
root@k8s-node-1:~# curl 172.31.12.2:8081
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver02, ServerIP: 172.31.12.12!
root@k8s-node-1:~# curl 172.31.12.2:8081
iKubernetes demoapp v1.0 !! ClientIP: 127.0.0.1, ServerName: webserver01, ServerIP: 172.31.12.11!
webserver01-app_1  | 127.0.0.1 - - [26/Sep/2022 01:47:12] "GET / HTTP/1.1" 200 -
webserver01-app_1  | 127.0.0.1 - - [26/Sep/2022 01:47:13] "GET / HTTP/1.1" 200 -
webserver02-app_1  | 127.0.0.1 - - [26/Sep/2022 01:47:14] "GET / HTTP/1.1" 200 -
webserver02-app_1  | 127.0.0.1 - - [26/Sep/2022 01:47:15] "GET / HTTP/1.1" 200 -
webserver01-app_1  | 127.0.0.1 - - [26/Sep/2022 01:47:16] "GET / HTTP/1.1" 200 -

猜你喜欢

转载自blog.csdn.net/qq_29974229/article/details/127082865
今日推荐
TIOBE 5 月榜单：Fortran “复活”进入 Top 10
GCC 14.1 发布
面壁智能发布 Eurux-8x22B 开源大模型 —— 堪称「理科状元」
开源日报 | 谷歌扶持鸿蒙上位；开源Rabbit R1；Docker加持的安卓手机；微软的焦虑和野心；海尔电器把开放平台关了
中国码农的“35岁魔咒”
蘭雅 CorelDRAW 插件 2024.5.1 国际劳动节版，免费下载
Arc Browser for Windows 1.0 正式 GA
90后程序员开发视频搬运软件、不到一年获利超 700 万，结局很刑！
周排行
Java自定义时间格式
同步整形电路
在开发中最最最常用的字符串的属性大集合
Linux 查看端口占用并杀掉
Java基础四：ArrayList
多线程之死锁就是这么简单
mysql 基础命令集
awk 命令详解
Centos6.3编译安装nginx+php步骤
OCR （Optical Character Recognition，光学字符识别）
每日归档
更多
2024-05-08(42)
2024-05-07(14)
2024-05-06(40)
2024-05-05(0)
2024-05-04(7)
2024-05-03(19)
2024-05-02(0)
2024-05-01(4)
2024-04-30(1)
2024-04-29(40)

代码天地 © 2018 codetd.com

境外服务器   最新电视剧2022