consul集群搭建

丹宇码农2025-12-01 16:10

一、安装

1.1.下载

consul和Nomad一样，都是一个独立的二进制文件，下载后解压，并将其拷贝到/usr/local/bin/目录下即可。

从 https://releases.hashicorp.com/consul 下载，目前最新的 1.22.0

bash 复制代码 
unzip consul_1.22.0_linux_amd64.zip
sudo mv consul /usr/local/bin/

1.2.端口规划

grpc_tls设置为-1，表示禁用gRPC TLS。

Server节点：

  • Server1: 8500 (HTTP), 8600 (DNS), 8300 (Server RPC), 8301 (Serf LAN), 8302 (Serf WAN)
  • Server2: 8501 (HTTP), 8601 (DNS), 8310 (Server RPC), 8311 (Serf LAN), 8312 (Serf WAN)
  • Server3: 8502 (HTTP), 8602 (DNS), 8320 (Server RPC), 8321 (Serf LAN), 8322 (Serf WAN)

Client节点：

  • Client1: 8503 (HTTP), 8603 (DNS), 8330 (Server RPC), 8331 (Serf LAN), 8332 (Serf WAN)
  • Client2: 8504 (HTTP), 8604 (DNS), 8340 (Server RPC), 8341 (Serf LAN), 8342 (Serf WAN)

注意：Serf WAN端口在Server节点上用于WAN gossip，在Client节点上通常不需要，但为了统一配置，我们都指定。

1.3.启动集群：

/data/consul/start-cluster.sh

停止所有consul进程：

复制代码 
pkill consul

1.4.验证集群状态：

检查集群成员

consul members -http-addr=http://127.0.0.1:8500

Node Address Status Type Build Protocol DC Partition Segment

server1 127.0.0.1:8301 alive server 1.22.0 2 dc1 default

server2 127.0.0.1:8311 alive server 1.22.0 2 dc1 default

server3 127.0.0.1:8321 alive server 1.22.0 2 dc1 default

client1 127.0.0.1:8331 alive client 1.22.0 2 dc1 default

client2 127.0.0.1:8341 alive client 1.22.0 2 dc1 default

检查Server节点状态

consul operator raft list-peers -http-addr=http://127.0.0.1:8500

检查节点服务

curl http://127.0.0.1:8500/v1/catalog/nodes

检查集群领导状态

curl http://127.0.0.1:8500/v1/status/leader

二、简单的http服务

见simple_app.py

复制代码 
安装依赖：
pip install Flask==2.3.3

conda activate qaanthing

指定端口（默认端口为5000）：
PORT=5000 python simple_app.py
PORT=5001 python simple_app.py

代码simple.py:

python 复制代码 
from flask import Flask, jsonify
import time
import socket
import os

app = Flask(__name__)

# 服务启动时间
start_time = time.time()

# 获取主机信息
def get_host_info():
    try:
        hostname = socket.gethostname()
        local_ip = socket.gethostbyname(hostname)
        return hostname, local_ip
    except:
        return 'unknown', 'unknown'

# 首页
@app.route('/')
def home():
    hostname, ip = get_host_info()
    return jsonify({
        'message': 'Hello from Simple Flask Service!',
        'service': 'simple-flask-service',
        'hostname': hostname,
        'ip': ip,
        'timestamp': time.strftime('%Y-%m-%d %H:%M:%S')
    })

# 健康检查接口 - Consul会使用这个
@app.route('/health')
def health():
    hostname, ip = get_host_info()
    return jsonify({
        'status': 'healthy',
        'service': 'simple-flask-service',
        'hostname': hostname,
        'ip': ip,
        'timestamp': time.strftime('%Y-%m-%d %H:%M:%S'),
        'uptime': round(time.time() - start_time, 2)
    }), 200

# 就绪检查
@app.route('/ready')
def ready():
    return jsonify({
        'status': 'ready',
        'timestamp': time.strftime('%Y-%m-%d %H:%M:%S')
    }), 200

# 存活检查
@app.route('/live')
def live():
    return jsonify({
        'status': 'alive',
        'timestamp': time.strftime('%Y-%m-%d %H:%M:%S')
    }), 200

# 服务信息
@app.route('/info')
def info():
    hostname, ip = get_host_info()
    return jsonify({
        'service': 'simple-flask-service',
        'version': '1.0.0',
        'hostname': hostname,
        'ip': ip,
        'python_version': os.environ.get('PYTHON_VERSION', 'unknown'),
        'environment': os.environ.get('ENVIRONMENT', 'development')
    })

if __name__ == '__main__':
    # 获取端口，默认5000
    port = int(os.environ.get('PORT', 5000))
    
    print(f"Starting Simple Flask Service on port {port}...")
    print(f"Health check available at: http://0.0.0.0:{port}/health")
    
    app.run(host='0.0.0.0', port=port, debug=False)

三、注册服务

仅用到标准的requests库。

python register_services.py

python 复制代码 
#!/usr/bin/env python3
import requests
import json
import time
import sys

class ConsulServiceRegistry:
    def __init__(self, consul_host='127.0.0.1', consul_port=8500):
        self.consul_base_url = f"http://{consul_host}:{consul_port}/v1"
    
    def register_service(self, service_name, service_id, address, port, tags=None, check=None):
        """
        注册服务到Consul
        """
        service_data = {
            "Name": service_name,
            "ID": service_id,
            "Address": address,
            "Port": port,
            "Tags": tags or []
        }
        
        # 添加健康检查配置
        if check:
            service_data["Check"] = check
        
        url = f"{self.consul_base_url}/agent/service/register"
        
        try:
            response = requests.put(url, json=service_data)
            if response.status_code == 200:
                print(f"✅ 服务注册成功: {service_name} (ID: {service_id})")
                return True
            else:
                print(f"❌ 服务注册失败: {service_name}, 状态码: {response.status_code}")
                return False
        except Exception as e:
            print(f"❌ 注册服务时出错: {service_name}, 错误: {str(e)}")
            return False
    
    def deregister_service(self, service_id):
        """
        从Consul注销服务
        """
        url = f"{self.consul_base_url}/agent/service/deregister/{service_id}"
        
        try:
            response = requests.put(url)
            if response.status_code == 200:
                print(f"✅ 服务注销成功: {service_id}")
                return True
            else:
                print(f"❌ 服务注销失败: {service_id}, 状态码: {response.status_code}")
                return False
        except Exception as e:
            print(f"❌ 注销服务时出错: {service_id}, 错误: {str(e)}")
            return False
    
    def list_services(self):
        """
        列出所有注册的服务
        """
        url = f"{self.consul_base_url}/agent/services"
        
        try:
            response = requests.get(url)
            if response.status_code == 200:
                services = response.json()
                print(f"📋 当前注册的服务 ({len(services)} 个):")
                for service_id, service_info in services.items():
                    print(f"  - {service_info['Service']} ({service_id}): {service_info['Address']}:{service_info['Port']}")
                return services
            else:
                print(f"❌ 获取服务列表失败, 状态码: {response.status_code}")
                return None
        except Exception as e:
            print(f"❌ 获取服务列表时出错: {str(e)}")
            return None

def main():
    # 创建Consul注册器
    consul = ConsulServiceRegistry('127.0.0.1', 8500)
    
    # 定义要注册的两个服务
    services_to_register = [
        {
            "name": "flask-app",
            "id": "flask-app-5000",
            "address": "127.0.0.1",
            "port": 5000,
            "tags": ["web", "python", "flask", "v1"],
            "check": {
                "name": "HTTP API Health Check",
                "http": f"http://127.0.0.1:5000/health",
                "method": "GET",
                "interval": "10s",
                "timeout": "5s"
            }
        },
        {
            "name": "flask-app", 
            "id": "flask-app-5001",
            "address": "127.0.0.1",
            "port": 5001,
            "tags": ["web", "python", "flask", "v1"],
            "check": {
                "name": "HTTP API Health Check",
                "http": f"http://127.0.0.1:5001/health",
                "method": "GET",
                "interval": "10s",
                "timeout": "5s"
            }
        }
    ]
    
    print("🚀 开始注册服务到Consul...")
    
    # 先列出当前服务
    print("\n--- 注册前的服务列表 ---")
    consul.list_services()
    
    # 注册服务
    print("\n--- 注册服务 ---")
    success_count = 0
    for service in services_to_register:
        if consul.register_service(
            service["name"],
            service["id"],
            service["address"],
            service["port"],
            service["tags"],
            service["check"]
        ):
            success_count += 1
    
    # 等待一下让注册生效
    time.sleep(2)
    
    # 显示注册后的服务列表
    print("\n--- 注册后的服务列表 ---")
    consul.list_services()
    
    print(f"\n📊 注册结果: {success_count}/{len(services_to_register)} 个服务注册成功")
    
    if success_count == len(services_to_register):
        print("🎉 所有服务注册成功!")
        
        # 显示健康检查信息
        print("\n🏥 健康检查端点:")
        for service in services_to_register:
            print(f"  - {service['id']}: {service['check']['http']}")
        
        print("\n🔍 你可以在Consul UI中查看服务状态: http://127.0.0.1:8500")
    else:
        print("⚠️  部分服务注册失败，请检查Consul状态和服务配置")

if __name__ == "__main__":
    main()

四、验证服务

检查Consul中所有服务的状态

仅用到标准的requests库。

python consul_check.py

python 复制代码 
#!/usr/bin/env python3
import requests
import json
from datetime import datetime

def check_consul_services(consul_host="127.0.0.1", consul_port=8500):
    """
    检查Consul中所有服务的状态
    """
    base_url = f"http://{consul_host}:{consul_port}/v1"
    
    print(f"🔍 检查Consul服务状态 - {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}")
    print(f"Consul地址: {consul_host}:{consul_port}")
    print("-" * 50)
    
    try:
        # 获取所有注册的服务
        services_url = f"{base_url}/agent/services"
        services_response = requests.get(services_url, timeout=5)
        
        if services_response.status_code != 200:
            print(f"❌ 无法获取服务列表: HTTP {services_response.status_code}")
            return
        
        services_data = services_response.json()
        
        if not services_data:
            print("ℹ️  Consul中没有注册的服务")
            return
        
        print(f"发现 {len(services_data)} 个注册的服务:")
        
        # 检查每个服务的健康状态
        for service_id, service_info in services_data.items():
            service_name = service_info['Service']
            address = service_info['Address']
            port = service_info['Port']
            
            # 获取服务的健康检查状态
            health_url = f"{base_url}/health/checks/{service_name}"
            health_response = requests.get(health_url, timeout=5)
            
            if health_response.status_code == 200:
                health_data = health_response.json()
                passing_checks = [check for check in health_data if check.get('Status') == 'passing']
                
                status = "✅ 健康" if len(passing_checks) > 0 else "❌ 不健康"
                check_count = f"{len(passing_checks)}/{len(health_data)}"
                
                print(f"  {status} {service_name} ({service_id})")
                print(f"    地址: {address}:{port}")
                print(f"    健康检查: {check_count}")
                
                # 显示标签（如果有）
                tags = service_info.get('Tags', [])
                if tags:
                    print(f"    标签: {', '.join(tags)}")
                print()
            else:
                print(f"  ⚠️  未知 {service_name} - 无法获取健康状态")
                print()
                
    except requests.exceptions.ConnectionError:
        print("❌ 无法连接到Consul，请确保Consul正在运行")
    except requests.exceptions.Timeout:
        print("❌ 连接Consul超时")
    except Exception as e:
        print(f"❌ 检查服务状态时出错: {e}")

def check_specific_service(service_name, consul_host="127.0.0.1", consul_port=8500):
    """
    检查特定服务的状态
    """
    base_url = f"http://{consul_host}:{consul_port}/v1"
    
    print(f"🔍 检查服务 '{service_name}' 的状态")
    print("-" * 50)
    
    try:
        # 获取服务的健康检查状态
        health_url = f"{base_url}/health/checks/{service_name}"
        health_response = requests.get(health_url, timeout=5)
        
        if health_response.status_code != 200:
            print(f"❌ 无法获取服务 '{service_name}' 的状态: HTTP {health_response.status_code}")
            return
        
        health_data = health_response.json()
        
        if not health_data:
            print(f"ℹ️  服务 '{service_name}' 没有健康检查或不存在")
            return
        
        # 按服务实例分组
        instances = {}
        for check in health_data:
            service_id = check.get('ServiceID', 'unknown')
            if service_id not in instances:
                instances[service_id] = {
                    'name': check.get('ServiceName', 'unknown'),
                    'address': check.get('ServiceAddress', 'unknown'),
                    'port': check.get('ServicePort', 'unknown'),
                    'checks': []
                }
            instances[service_id]['checks'].append({
                'name': check.get('Name', 'unknown'),
                'status': check.get('Status', 'unknown'),
                'output': check.get('Output', '')
            })
        
        print(f"服务 '{service_name}' 有 {len(instances)} 个实例:")
        
        for instance_id, instance_info in instances.items():
            passing_checks = [c for c in instance_info['checks'] if c['status'] == 'passing']
            status = "✅ 健康" if len(passing_checks) == len(instance_info['checks']) else "❌ 不健康"
            
            print(f"  {status} {instance_id}")
            print(f"    地址: {instance_info['address']}:{instance_info['port']}")
            print(f"    检查结果: {len(passing_checks)}/{len(instance_info['checks'])} 通过")
            
            # 显示失败的检查详情
            for check in instance_info['checks']:
                icon = "✅" if check['status'] == 'passing' else "❌"
                print(f"    {icon} {check['name']}: {check['status']}")
                if check['status'] != 'passing' and check['output']:
                    print(f"        详情: {check['output']}")
            print()
            
    except Exception as e:
        print(f"❌ 检查服务 '{service_name}' 时出错: {e}")

def get_healthy_instances(service_name, consul_host="127.0.0.1", consul_port=8500):
    """
    获取特定服务的所有健康实例
    """
    base_url = f"http://{consul_host}:{consul_port}/v1"
    
    try:
        # 使用健康端点获取健康的服务实例
        health_url = f"{base_url}/health/service/{service_name}?passing=true"
        response = requests.get(health_url, timeout=5)
        
        if response.status_code != 200:
            print(f"❌ 无法获取健康实例: HTTP {response.status_code}")
            return []
        
        instances_data = response.json()
        healthy_instances = []
        
        for instance in instances_data:
            service_info = instance['Service']
            healthy_instances.append({
                'id': service_info['ID'],
                'name': service_info['Service'],
                'address': service_info['Address'],
                'port': service_info['Port'],
                'tags': service_info.get('Tags', [])
            })
        
        return healthy_instances
        
    except Exception as e:
        print(f"❌ 获取健康实例时出错: {e}")
        return []

if __name__ == "__main__":
    # 检查所有服务状态
    check_consul_services()
    
    print("\n" + "="*50)
    print("详细服务检查")
    print("="*50)
    
    # 检查特定服务
    check_specific_service("flask-app")
    
    # 获取健康的服务实例
    healthy_instances = get_healthy_instances("flask-app")
    if healthy_instances:
        print(f"🎯 发现 {len(healthy_instances)} 个健康的 'flask-app' 实例:")
        for instance in healthy_instances:
            print(f"  - {instance['id']}: {instance['address']}:{instance['port']}")
    else:
        print("ℹ️  没有发现健康的 'flask-app' 实例")

会有类似如下输出：

服务 'flask-app' 有 2 个实例:

✅ 健康 flask-app-5000

地址: unknown:unknown

检查结果: 1/1 通过

✅ HTTP API Health Check: passing

❌ 不健康 flask-app-5001

地址: unknown:unknown

检查结果: 0/1 通过

❌ HTTP API Health Check: critical

详情: Get "http://127.0.0.1:5001/health": dial tcp 127.0.0.1:5001: connect: connection refused

相关推荐
努力的小帅2 小时前
Python_OpenCV入门到精通——入门篇（看这一篇就足够了！！！）
图像处理·python·opencv·计算机视觉·视觉检测·视觉项目
洗紫2 小时前
Python常用内置模块教程：os/sys/datetime/random/json全解析
python
z***89712 小时前
Flask框架中SQLAlchemy的使用方法
后端·python·flask
G果2 小时前
ROS2 Cartographer纯定位导航遇到的问题
python·ros2·定位·cartographer·导航·launch·navigation2
Brduino脑机接口技术答疑3 小时前
脑机接口数据处理连载（二） 数据源头解析：脑电信号的采集原理与数据特性
人工智能·python·算法·数据分析·脑机接口
Wise玩转AI3 小时前
Day 25｜智能体的“可观察性（Observability）体系”
人工智能·python·ai·chatgpt·ai智能体
p***92483 小时前
Python使用PyMySQL操作MySQL完整指南
数据库·python·mysql
kali-Myon3 小时前
NewStarCTF2025-Week5-Web
java·python·安全·web安全·php·web·ctf
DFT计算杂谈3 小时前
Abinit-10.4.7安装教程
linux·数据库·python·算法·matlab
热门推荐
01GitHub 镜像站点02BongoCat - 跨平台键盘猫动画工具03UV安装并设置国内源04Linux下V2Ray安装配置指南05安娜的档案(Anna’s Archive) 镜像网站/国内最新可访问入口（持续更新）06【保姆级教程】免费使用Gemini3的5种方法！免翻墙/国内直连0746个Nano-banana 精选提示词，持续更新中08本地部署阿里最新开源的Z-Image09Meta第三代“分割一切”模型——SAM 3本地部署教程：首支持文本提示分割，400万概念、30毫秒响应，检测分割追踪一网打尽10在VSCode配置Java开发环境的保姆级教程（适配各类AI编程IDE）