Docker容器内Flask API无法调用另一容器REST API的问题排查

通过Docker Compose分别部署了Flask API（hardware-reservation-api）+ MongoDB容器，以及Go语言开发的REST API（config-manager）。Flask API尝试调用http://config-manager:4000/env-config时出现超时，但本地运行Flask API并改用localhost:4000调用时可正常工作。

相关配置如下：

Flask服务的docker-compose.yaml #

version: '3.9'
services:
  mongo:
    image: mongo
    ports:
      - "27017:27017"
  hardware-reservation-api:
    build: .
    ports:
      - "7070:7070"
    environment:
      MONGO_URI: mongodb://mongo:27017/
    depends_on:
      - mongo

Flask的Dockerfile #

FROM python:3.6.5-slim

WORKDIR /app

ADD . /app

RUN pip install --trusted-host pypi.python.org -r requirements.txt

EXPOSE 7070

CMD ["python", "app.py"]

config-manager的docker-compose.yaml #

version: '3.9'
services:
  config-manager:
    build: .
    container_name: config-manager
    ports:
      - "4000:4000"

config-manager的Dockerfile #

FROM golang:latest

# Set the working directory to /app
WORKDIR /app

# Copy the go.mod and go.sum files to the container
COPY go.mod .
COPY go.sum .

# Download all dependencies. This step can be cached unless the mod/sum files are changed.
RUN go mod download

# Copy the current directory contents into the container at /app
COPY . .

# Build the Go app
RUN go build -o main .

# Expose port 4000 for the API to listen on
EXPOSE 4000

CMD ["./main"]

Flask API调用代码 #

api_url = config.get('flask', 'config_manager_url', fallback='http://config-manager:4000/env-config')
params = {'path': environment_name}
response = requests.get(api_url, params=params)

if response.status_code == 200:
# The API call was successful
    response_data = yaml.load(response.content, Loader=yaml.FullLoader)
    return response_data, ""
else:
    val = f"Error: fetching configuration failed from config manager {response.status_code} - {response.text}"
    return None, val

Docker Compose默认会为每个独立的docker-compose.yaml文件创建专属默认网络，两个不同的Compose项目处于完全隔离的网络环境中。Flask容器所在的网络无法解析config-manager这个服务名称，也无法访问另一个网络内的容器，因此出现调用超时。

方法1：合并到同一个Docker Compose文件 #

将Flask、MongoDB和config-manager服务放到同一个docker-compose.yaml中，所有服务会自动加入同一默认网络，服务名可直接相互解析：

version: '3.9'
services:
  mongo:
    image: mongo
    ports:
      - "27017:27017"
  hardware-reservation-api:
    build: ./path-to-flask-project # 调整为Flask项目的实际目录路径
    ports:
      - "7070:7070"
    environment:
      MONGO_URI: mongodb://mongo:27017/
    depends_on:
      - mongo
      - config-manager
  config-manager:
    build: ./path-to-go-project # 调整为Go项目的实际目录路径
    ports:
      - "4000:4000"

方法2：创建共享自定义网络 #

手动创建一个Docker网络，让两个Compose项目都加入该网络，实现跨项目服务通信：

1. 创建自定义网络：

docker network create shared-services-network

2. 修改Flask的docker-compose.yaml，加入共享网络：

version: '3.9'
services:
  mongo:
    image: mongo
    ports:
      - "27017:27017"
    networks:
      - shared-network
  hardware-reservation-api:
    build: .
    ports:
      - "7070:7070"
    environment:
      MONGO_URI: mongodb://mongo:27017/
    depends_on:
      - mongo
    networks:
      - shared-network
networks:
  shared-network:
    external: true
    name: shared-services-network

3. 修改config-manager的docker-compose.yaml，加入共享网络：

version: '3.9'
services:
  config-manager:
    build: .
    container_name: config-manager
    ports:
      - "4000:4000"
    networks:
      - shared-network
networks:
  shared-network:
    external: true
    name: shared-services-network

方法3：使用宿主机IP访问（不推荐） #

让Flask调用宿主机的实际IP（例如http://192.168.x.x:4000/env-config），但该方法依赖宿主机网络配置，移植性差，不适合生产环境。

内容的提问来源于stack exchange，提问作者Basu