Docker容器下Spring Boot+Testcontainers+Gradle集成测试失败

问题概述 #

为Kafka生产者消费者应用搭建了集成测试环境，本地IDE运行Gradle测试完全正常，但在Docker驱动的CI/CD流水线中，Testcontainers无法自动拉取Kafka镜像，测试失败并抛出以下错误：

environment initializationError FAILED 
java.lang.IllegalStateException at DockerClientProviderStrategy.java

现有配置详情 #

1. 集成测试代码 #

@Testcontainer
@DirtiesContext
@SpringBootTest
@ExtendWith(SpringExtension.class)
public class KafkaIntegrationTest {
        
   @Container
   private static final KafkaContainer kafkaContainer = new KafkaContainer(DockerImageName.parse("confluentinc/cp-kafka:latest"))
          .withFileSystemBind("/var/run/docker.sock", "/var/run/docker.sock", BindMode.READ_WRITE);
    
   @DynamicPropertySource
   private static void kafkaProperties(DynamicPropertyRegistry registry) {
      registry.add("spring.kafka.bootstrap-servers", KafkaContainer::getBootstrapServers);
   }
        
   @Test
   public void givenKafkaContainerWhenSendingWithProducerThenMessageReceived() {
      // Consumer integration test here
   }
                
}

2. Dockerfile配置 #

...
FROM docker:latest
...
VOLUME /var/run/docker.sock:/var/run/docker.sock
...
RUN ./gradlew --no-daemon testClasses processIntegrationTests
...

3. Jenkinsfile配置 #

...
    agent {
        docker {
            image 'docker:latest'
            args '-v /var/run/docker.sock:/var/run/docker.sock --privileged'
        }
    }
    ...
    stage('Test') {
        steps {
            sh "docker build --target test-result ."
        }
    ...
    }

问题原因及解决方法 #

核心原因 #

1. docker:latest镜像默认未安装Java环境，Testcontainers依赖Java运行，无法正常初始化
2. Docker套接字挂载的权限配置错误，容器内进程无权限访问宿主机Docker资源
3. Dockerfile中用VOLUME声明挂载属于构建时配置，无法正确传递宿主机的套接字权限

分步解决方案 #

1. 修正Dockerfile配置

替换基础镜像并补充必要依赖，移除无效的VOLUME声明：

# 使用带Java的基础镜像，同时安装Docker客户端
FROM openjdk:17-jdk-slim

# 安装Docker客户端及依赖
RUN apt-get update && apt-get install -y docker.io && rm -rf /var/lib/apt/lists/*

# 调整Docker套接字权限，确保容器内进程可访问
RUN chmod 666 /var/run/docker.sock

# 复制项目代码到容器内
COPY . .

# 执行测试前置任务
RUN ./gradlew --no-daemon testClasses processIntegrationTests

2. 调整Jenkinsfile配置

优化代理镜像配置，提前拉取Kafka镜像避免超时：

...
    agent {
        docker {
            image 'openjdk:17-jdk-slim'
            args '-v /var/run/docker.sock:/var/run/docker.sock -v $(which docker):/usr/bin/docker --privileged'
        }
    }
    ...
    stage('预拉取Kafka镜像') {
        steps {
            sh "docker pull confluentinc/cp-kafka:latest"
        }
    }
    stage('执行测试') {
        steps {
            sh "./gradlew --no-daemon integrationTest"
        }
    ...
    }

3. 简化测试代码配置

移除测试类中多余的套接字绑定配置，由CI容器层面统一处理：

@Testcontainer
@DirtiesContext
@SpringBootTest
@ExtendWith(SpringExtension.class)
public class KafkaIntegrationTest {
        
   @Container
   private static final KafkaContainer kafkaContainer = new KafkaContainer(DockerImageName.parse("confluentinc/cp-kafka:latest"));
    
   @DynamicPropertySource
   private static void kafkaProperties(DynamicPropertyRegistry registry) {
      registry.add("spring.kafka.bootstrap-servers", kafkaContainer::getBootstrapServers);
   }
        
   @Test
   public void givenKafkaContainerWhenSendingWithProducerThenMessageReceived() {
      // Consumer integration test here
   }
                
}

关键注意事项 #

• 禁止在Dockerfile中用VOLUME声明Docker套接字挂载，该指令仅用于定义容器数据卷，不适用于宿主机资源挂载
• 确保CI容器内同时具备Java运行环境和Docker客户端，二者是Testcontainers正常工作的必要条件
• 提前拉取镜像可避免网络波动导致的拉取超时问题
• 运行CI容器时必须添加--privileged参数，确保容器拥有足够权限访问宿主机Docker资源

内容的提问来源于stack exchange，提问作者Dave