环境准备
Nacos: v2.3.0
Spring Boot: 3.2.0
Spring Cloud: 2023.0.0
Spring Cloud Alibaba Version: 2023.0.0.0-RC1
项目编写
github 地址：https://github.com/yuluo-yx/sca-k8s-demo/tree/openfeign
项目结构
├─docker-compose # Docker compose 部署文件(这里只是示例，没有任何其他用途)
├─kubernetes # Kubernetes 部署文件
└─docker-images
├─consumer
├─ application-k8s.yaml # k8s 环境的配置文件
├─ app.jar # 应用 jar 包
└─ Dockerfile # 打包的 Dockerfile
└─provider
├─sca-k8s-service-consumer # sca 服务消费者模块
├─sca-k8s-service-provider # sca 服务提供者模块
父 pom.xml 文件：
<modules>
<module>sca-k8s-service-provider</module>
<module>sca-k8s-service-consumer</module>
</modules>
<properties>
<maven.compiler.source>17</maven.compiler.source>
<maven.compiler.target>17</maven.compiler.target>
<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
<spring-cloud-alibaba.version>2023.0.0.0-RC1</spring-cloud-alibaba.version>
<spring-cloud.version>2023.0.0</spring-cloud.version>
<spring-boot.version>3.2.0</spring-boot.version>
</properties>
<dependencies>
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-web</artifactId>
</dependency>
<dependency>
<groupId>com.alibaba.cloud</groupId>
<artifactId>spring-cloud-starter-alibaba-nacos-discovery</artifactId>
</dependency>
</dependencies>
<dependencyManagement>
<dependencies>
<dependency>
<groupId>org.springframework.cloud</groupId>
<artifactId>spring-cloud-dependencies</artifactId>
<version>${spring-cloud.version}</version>
<type>pom</type>
<scope>import</scope>
</dependency>
<dependency>
<groupId>com.alibaba.cloud</groupId>
<artifactId>spring-cloud-alibaba-dependencies</artifactId>
<version>${spring-cloud-alibaba.version}</version>
<type>pom</type>
<scope>import</scope>
</dependency>
</dependencies>
</dependencyManagement>
<!--spring boot 打包插件-->
<build>
<plugins>
<plugin>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-maven-plugin</artifactId>
</plugin>
</plugins>
</build>
provider 模块
provider pom.xml 文件
<dependencies>
<dependency>
<groupId>com.alibaba.cloud</groupId>
<artifactId>spring-cloud-starter-alibaba-nacos-discovery</artifactId>
</dependency>
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-actuator</artifactId>
</dependency>
</dependencies>
provider controller java 文件
@RestController
@RequestMapping("/provider")
public class ProviderController {
@Resource
private ProviderService providerService;
@GetMapping("/a")
public String providerA() {
return providerService.providerA();
}
@GetMapping("/b")
public String providerB() {
return providerService.providerB();
}
}
provider service java 文件
@Service
public class ProviderServiceImpl implements ProviderService {
@Override
public String providerA() {
return "This response from provider A!";
}
@Override
public String providerB() {
return "This response from provider B!";
}
}
provider application.yml 配置
server:
port: 8082
spring:
application:
name: sca-k8s-provider
cloud:
nacos:
discovery:
# nacos 服务发现地址
server-addr: 127.0.0.1:8848
username: 'nacos'
password: 'nacos'
# actuator 健康检查配置，用于做 kubernetes pod 的 liveness 探针接口使用
management:
server:
# management 类端口全部保持一致 （尽量让业务服务的 K8S Deployment 保持统一，便于维护）
port: 30000
endpoint:
health:
probes:
enabled: true
endpoints:
web:
exposure:
include: health
consumer 模块
Consumer 模块使用
spring-cloud-openfeign作为 rpc 组件。
consumer pom.xml 文件
<dependencies>
<dependency>
<groupId>com.alibaba.cloud</groupId>
<artifactId>spring-cloud-starter-alibaba-nacos-discovery</artifactId>
</dependency>
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-actuator</artifactId>
</dependency>
<dependency>
<groupId>org.springframework.cloud</groupId>
<artifactId>spring-cloud-starter-openfeign</artifactId>
</dependency>
<dependency>
<groupId>org.springframework.cloud</groupId>
<artifactId>spring-cloud-starter-loadbalancer</artifactId>
</dependency>
<dependency>
<groupId>com.alibaba.cloud</groupId>
<artifactId>spring-cloud-starter-alibaba-sentinel</artifactId>
</dependency>
</dependencies>
consumer feign client java 文件
@FeignClient(
name = "sca-k8s-provider",
fallback = K8sFeignCallback.class,
configuration = K8sFeignConfig.class,
contextId = "sca-k8s-provider"
)
public interface K8sFeignClient {
@GetMapping("/provider/a")
String providerA();
@GetMapping("/provider/b")
String providerB();
}
consumer feign config java 文件
public class K8sFeignConfig {
@Bean
public K8sFeignCallback feignCallback() {
return new K8sFeignCallback();
}
}
Consumer feign fallback java 文件
public class K8sFeignCallback implements K8sFeignClient{
@Override
public String providerA() {
return "call provider A interface failed, fallback";
}
@Override
public String providerB() {
return "call provider B interface failed, fallback";
}
}
Consumer controller java
@RestController
@RequestMapping("/consumer")
public class ConsumerController {
@Autowired
private ConsumerService consumerService;
@GetMapping("/a")
public String consumerA() {
return consumerService.consumerA();
}
@GetMapping("/b")
public String consumerB() {
return consumerService.consumerB();
}
}
Consumer service java 文件
@Service
public class ConsumerServiceImpl implements ConsumerService {
@Autowired
private K8sFeignClient feignClient;
@Override
public String consumerA() {
return feignClient.providerA();
}
@Override
public String consumerB() {
return feignClient.providerB();
}
}
Consumer application 主类 java 文件
@EnableFeignClients
@EnableDiscoveryClient
@SpringBootApplication
@LoadBalancerClients({
@LoadBalancerClient("sca-k8s-provider")
})
public class SCAK8sConsumerApplication {
public static void main(String[] args) {
SpringApplication.run(SCAK8sConsumerApplication.class, args);
}
}
consumer application.yml 配置
server:
port: 8080
spring:
application:
name: sca-k8s-consumer
cloud:
nacos:
discovery:
server-addr: 127.0.0.1:8848
username: 'nacos'
password: 'nacos'
management:
server:
# management 类端口全部保持一致 （尽量让业务服务的 K8S Deployment 保持统一，便于维护）
port: 30000
endpoint:
health:
probes:
enabled: true
endpoints:
web:
exposure:
include: health
feign:
sentinel:
enabled: true
部署效果展示
本项目提供了 Docker compose 和 Kubernetes 部署两种方式。
Docker compose 运行方式参考
docker-compose目录下的 README.md 文件。本节中主要以 k8s 部署方式演示为主。本地运行需要准备 2.3.0 版本的 nacos server。
环境准备
准备 k8s 集群，此处为了演示方便，使用 kind 模拟 k8s cluster，初始化集群时创建的 pod 如下所示：
root@yuluo-Inspiron-3647:/kubernetes/sca-k8s-best# kubectl get pods -A
NAMESPACE NAME READY STATUS RESTARTS AGE
kube-system coredns-76f75df574-74qnn 1/1 Running 0 118s
kube-system coredns-76f75df574-q9hq5 1/1 Running 0 118s
kube-system etcd-sca-k8s-control-plane 1/1 Running 0 2m19s
kube-system kindnet-bmsmm 1/1 Running 0 118s
kube-system kube-apiserver-sca-k8s-control-plane 1/1 Running 0 2m24s
kube-system kube-controller-manager-sca-k8s-control-plane 1/1 Running 0 2m19s
kube-system kube-proxy-lllzb 1/1 Running 0 118s
kube-system kube-scheduler-sca-k8s-control-plane 1/1 Running 0 2m23s
local-path-storage local-path-provisioner-7577fdbbfb-zpwb4 1/1 Running 0 117s
部署流程
部署单节点 Nacos
执行 kubectl create -f sca-k8s-demo-mysql.yaml 创建 nacos 需要的 mysql 服务：
root@yuluo-Inspiron-3647:/kubernetes/sca-k8s-best# kubectl create -f sca-k8s-demo-mysql.yaml
replicationcontroller/sca-k8s-demo-mysql created
service/sca-k8s-demo-mysql created
创建成功使用 kubectl get pods -A | grep mysql 查询，如下所示：
root@yuluo-Inspiron-3647:/kubernetes/sca-k8s-best# kubectl get pods -A | grep mysql
default sca-k8s-demo-mysql-87wft 1/1 Running 0 2m5s
执行 kubectl create -f sca-k8s-demo-nacos.yaml 创建 nacos 需要的 mysql 服务：
root@yuluo-Inspiron-3647:/kubernetes/sca-k8s-best# kubectl create -f sca-k8s-demo-nacos.yaml
service/sca-k8s-demo-nacos-standalone created
configmap/sca-k8s-demo-nacos-cm created
deployment.apps/sca-k8s-demo-nacos-standalone created
创建成功如下所示：
root@yuluo-Inspiron-3647:/kubernetes/sca-k8s-best# kubectl get pods | grep nacos
sca-k8s-demo-nacos-standalone-854d8cfc88-rg7pp 1/1 Running 0 2m26s
这里没有使用 ingress 暴露 nacos 服务，使用端口转发的方式将 nacos 暴露出来。
nacos 被设计成在内网使用的中间件，在生产环境中应注意网络隔离策略。
暴露 nacos server：kubectl port-forward --address localhost,192.168.20.129 svc/sca-k8s-demo-nacos-standalone 8848:8848 9848:9848
这里的地址是虚拟机的 ip 地址，实际操作时注意 ip 地址的变化。
浏览器访问 nacos console：
打包 Docker Images
修改应用配置文件中的 nacos server 地址为端口转发时使用的虚拟机地址，确保服务成功注册。consumer 和 provider 同时修改。
为了使用方便，此处使用阿里云容器镜像服务。执行下述命令之前先使用 docker login 命令登陆 docker hub，确保 push 镜像成功！
修改 docker images 的 tag 为自己的阿里云镜像仓库用户名地址。之后执行 ./kubernetes/docker-images/build.sh 脚本
部署 Provider 和 Consumer 服务
k8s 部署资源文件解析
ConfigMap 资源文件：用于设置一些 jvm 调优参数和指定激活的配置文件：
更最佳的实践做法为：线上的配置不会写在 dockerfile 或者其他构件相关的配置文件中，而是直接维护在 k8s 的 ConfigMap 里面 （通常由运维人员维护，开发者不关注这些具体的参数信息）
通常情况下：应用的配置文件为
application.yml # 通用配置项
application-dev.yml # 开发环境配置
application-test.yml # 测试环境配置
application-prod.yml # 生产环境配置在此 demo 设置应用配置文件为 k8s，如上文项目结构所示！
apiVersion: v1
data:
spring-profiles-active: 'k8s'
kind: ConfigMap
metadata:
name: sca-k8s-demo-spring-profile-cm
provider deployment 资源文件：
apiVersion: apps/v1
kind: Deployment
metadata:
labels:
app: sca-k8s-demo-provider-service
name: sca-k8s-demo-provider-service
spec:
replicas: 1
selector:
matchLabels:
app: sca-k8s-demo-provider-service
template:
metadata:
labels:
app: sca-k8s-demo-provider-service
spec:
containers:
- env:
- name: SPRING_PROFILES_ACTIVE
valueFrom:
configMapKeyRef:
name: sca-k8s-demo-spring-profile-cm
key: spring-profiles-active
# jvm arguments
# - name: JAVA_OPTION
# value: "-Dfile.encoding=UTF-8 -XX:+UseParallelGC -XX:+PrintGCDetails -Xloggc:/var/log/devops-example.gc.log -XX:+HeapDumpOnOutOfMemoryError -XX:+DisableExplicitGC"
# - name: XMX
# value: "128m"
# - name: XMS
# value: "128m"
# - name: XMN
# value: "64m"
name: sca-k8s-demo-provider-service
image: registry.cn-hangzhou.aliyuncs.com/yuluo-yx/sca-k8s-demo-provider-service:latest
ports:
- containerPort: 8082
livenessProbe:
httpGet:
path: /actuator/health/liveness
# port 尽量保持统一，方便维护
port: 30000
scheme: HTTP
initialDelaySeconds: 20
periodSeconds: 10
readinessProbe:
httpGet:
path: /actuator/health/readiness
port: 30000
scheme: HTTP
initialDelaySeconds: 20
periodSeconds: 10
consumer 对比 provider 多了一个 svc 配置，其他相同。之后需要做端口转发，提供外部访问，同样也可以使用 ingress 暴露服务出去。
部署
修改完成所有配置之后，执行 kubectl create -f sca-k8s-demo-cm.yaml 部署 provider 和 consumer 需要的 ConfigMap 资源，
之后部署 provider service kukbectl create -f sca-k8s-demo-provider.yaml：
root@yuluo-Inspiron-3647:/kubernetes/sca-k8s-best# kubectl get pods -A | grep provider
default sca-k8s-demo-provider-service-7c6579956b-xswjq 1/1 Running 0 9m12s
执行 kubectl logs sca-k8s-demo-provider-service-xxxx 查看 pod 日志，发现激活的配置文件为 k8s，证明 ConfigMap 生效：
查看 nacos 控制台，发现 provider 已经注册：
尝试访问 provider service 接口服务：
转发 provider service pod：kubectl port-forward pod-name 8082:8082
root@yuluo-Inspiron-3647:/kubernetes/sca-k8s-best# curl 127.0.0.1:8082/provider/a
This response from provider A!
consumer 部署方式相同，执行 kubectl create -f sca-k8s-demo-consumer.yaml 即可部署。
部署之后查看 nacos 注册中心：
最终部署所有的 k8s 资源如下：
root@yuluo-Inspiron-3647:/kubernetes/sca-k8s-best# kubectl get all | grep sca
pod/sca-k8s-demo-consumer-service-7ffff75969-zmtcw 1/1 Running 0 2m5s
pod/sca-k8s-demo-mysql-zmstn 1/1 Running 0 35m
pod/sca-k8s-demo-nacos-standalone-854d8cfc88-rg7pp 1/1 Running 0 34m
pod/sca-k8s-demo-provider-service-7c6579956b-xswjq 1/1 Running 0 15m
replicationcontroller/sca-k8s-demo-mysql 1 1 1 35m
service/sca-k8s-demo-consumer-service-svc ClusterIP 10.96.61.65 <none> 8080/TCP 2m5s
service/sca-k8s-demo-mysql ClusterIP 10.96.37.255 <none> 3306/TCP 35m
service/sca-k8s-demo-nacos-standalone ClusterIP None <none> 8848/TCP,9848/TCP,9849/TCP,7848/TCP 34m
deployment.apps/sca-k8s-demo-consumer-service 0/1 1 0 2m5s
deployment.apps/sca-k8s-demo-nacos-standalone 1/1 1 1 34m
deployment.apps/sca-k8s-demo-provider-service 1/1 1 1 15m
replicaset.apps/sca-k8s-demo-consumer-service-7ffff75969 1 1 0 2m5s
replicaset.apps/sca-k8s-demo-nacos-standalone-854d8cfc88 1 1 1 34m
replicaset.apps/sca-k8s-demo-provider-service-7c6579956b 1 1 1 15m
访问
转发 consumer 服务：kubectl port-forward --address localhost,192.168.20.129 svc/sca-k8s-demo-consumer-service-svc 8080:8080
root@yuluo-Inspiron-3647:/kubernetes/sca-k8s-best# curl 127.0.0.1:8080/consumer/a
This response from provider A!
root@yuluo-Inspiron-3647:/kubernetes/sca-k8s-best# curl 127.0.0.1:8080/consumer/b
This response from provider B!
释放资源
kubectl delete -f sca-k8s-demo-*.yaml
本文章主要介绍如何在 Kubernetes 环境中部署 Spring Cloud Alibaba 应用。在部署的同时，使用 liveness 探针确保 pod 正常启动可对外提供服务，使用 ConfigMap 配置使得应用的配置参数更加灵活。