Spring Boot+Docker环境下Actuator/Prometheus请求超时问题排查

执行docker-compose up -d后所有服务启动正常，本地访问localhost:1221/actuator/prometheus可正常获取如下指标数据：

# HELP jvm_buffer_count_buffers An estimate of the number of buffers in the pool
# TYPE jvm_buffer_count_buffers gauge
jvm_buffer_count_buffers{id="mapped - 'non-volatile memory'",} 0.0
jvm_buffer_count_buffers{id="mapped",} 0.0
jvm_buffer_count_buffers{id="direct",} 9.0
# HELP spring_security_filterchains_requestcache_before_total  
# TYPE spring_security_filterchains_requestcache_before_total counter
spring_security_filterchains_requestcache_before_total{security_security_reached_filter_section="before",spring_security_filterchain_position="0",spring_security_filterchain_size="0",spring_security_reached_filter_name="none",} 7.0
....

但访问http://localhost:9090/targets时，Prometheus提示如下错误：

Get "http://book-delivery-app-1:1221/actuator/prometheus": context deadline exceeded

相关配置文件 #

docker-compose.yml

version: "3.9"

services:
  database:
    container_name: database
    image: mysql:8.0.33
    restart: always
    env_file:
      - .env  # 使用.env文件存储环境变量
    environment:
      MYSQL_DATABASE: bookdelivery
      MYSQL_PASSWORD: ${DATABASE_PASSWORD}
      MYSQL_ROOT_PASSWORD: ${DATABASE_PASSWORD}
      MYSQL_ROOT_HOST: '%'
      MYSQL_PORT: 3307
    volumes:
      - ./db:/var/lib/mysql
    ports:
      - "3307:3306"
    networks:
      - bookDeliveryNetwork

  bookdelivery:
    image: 'bookdelivery:latest'
    build:
      context: .
      dockerfile: Dockerfile
    container_name: bookdelivery
    restart: on-failure
    env_file:
      - .env  # 使用.env文件存储环境变量
    ports:
      - "1221:1221"
    environment:
      - server.port=1221
      - spring.datasource.username=${DATABASE_USERNAME}
      - spring.datasource.password=${DATABASE_PASSWORD}
      - BOOK_DELIVERY_DB_IP=database
      - BOOK_DELIVERY_DB_PORT=3307
      - spring.datasource.url=jdbc:mysql://host.docker.internal:3307/bookdelivery
    depends_on:
      - database
    networks:
      - bookDeliveryNetwork

  prometheus:
    image: prom/prometheus:v2.35.0
    container_name: prometheus
    restart: unless-stopped
    ports:
      - "9090:9090"
    volumes:
      - ./data/prometheus/config:/etc/prometheus/
    command:
      - '--config.file=/etc/prometheus/prometheus.yml'
    networks:
      - bookDeliveryNetwork

  grafana:
    image: grafana/grafana-oss:8.5.2
    pull_policy: always
    container_name: grafana
    restart: unless-stopped
    ports:
      - "3000:3000"
    volumes:
      - ./data/grafana:/var/lib/grafana
    environment:
      - GF_SECURITY_ADMIN_PASSWORD=admin
      - GF_SERVER_DOMAIN=localhost
    networks:
      - bookDeliveryNetwork


networks:
  bookDeliveryNetwork:

prometheus.yml

global:
  scrape_interval: 120s # 默认每15秒抓取一次目标，此处修改为120秒
  evaluation_interval: 120s # 默认每15秒评估一次规则，此处修改为120秒

# 抓取配置，包含一个要抓取的端点：
# 这里是Prometheus自身
scrape_configs:
  # 作业名称会作为标签`job=<job_name>`添加到从此配置抓取的所有时间序列中
  - job_name: 'prometheus'
    # 覆盖全局默认，每5秒抓取一次此作业的目标
    scrape_interval: 5s
    # metrics_path默认为'/metrics'
    # scheme默认为'http'
    static_configs:
      - targets: ['localhost:9090']

  - job_name: 'Spring Boot Application input'
    metrics_path: '/actuator/prometheus'
    scrape_interval: 2s
    static_configs:
      - targets: [ 'book-delivery-app-1:1221' ]
        labels:
          application: 'Book Delivery API'

问题根源与修复方案 #

1. 核心问题：Prometheus目标地址错误 #

在Docker自定义网络bookDeliveryNetwork中，服务间通信必须使用服务名称或容器名称作为访问地址，而非Docker自动生成的容器前缀名book-delivery-app-1。你的Spring Boot服务在docker-compose中定义的container_name是bookdelivery，服务名也是bookdelivery，Prometheus无法解析不存在的book-delivery-app-1地址，导致请求超时。

2. 修复步骤 #

步骤1：修正Prometheus配置

修改prometheus.yml中的Spring Boot抓取目标，将：

targets: [ 'book-delivery-app-1:1221' ]

改为：

targets: [ 'bookdelivery:1221' ]

步骤2：优化Spring Boot数据库连接配置（可选但推荐）

当前bookdelivery服务的spring.datasource.url使用了host.docker.internal:3307，在自定义Docker网络中，服务间应直接通过服务名访问容器内部端口，而非主机映射端口。将其改为：

spring.datasource.url=jdbc:mysql://database:3306/bookdelivery

此修改可避免依赖主机端口映射，让服务间通信更稳定。

步骤3：重启服务

执行以下命令重启Prometheus（或整个docker-compose服务栈）：

# 仅重启Prometheus
docker-compose restart prometheus

# 或重启所有服务
docker-compose down && docker-compose up -d

内容的提问来源于stack exchange，提问作者Sercan Noyan Germiyanoğlu