当前位置：首页 > news >正文

系统架构：高可用与容错设计

news 2026/6/4 7:44:32

系统架构：高可用与容错设计

大家好，我是欧阳瑞（Rich Own）。今天想和大家聊聊系统架构这个重要话题。作为一个全栈开发者，高可用和容错设计是构建可靠系统的关键。今天就来分享一下高可用与容错设计的实战经验。

高可用概述

什么是高可用性？

高可用性是指系统在长时间内保持正常运行的能力 通常用几个9来衡量：99.9%、99.99%、99.999%

可用性计算

可用性 = (总时间 - 停机时间) / 总时间 × 100% 99.9% → 每年停机约8.77小时 99.99% → 每年停机约52.6分钟 99.999% → 每年停机约5.26分钟

核心策略

冗余设计 → 多副本部署 故障转移 → 自动切换 负载均衡 → 流量分发 监控告警 → 及时发现问题

容错设计

故障处理策略

策略	说明	适用场景
重试	失败后重新请求	网络抖动、临时故障
熔断	暂时停止请求	服务持续故障
降级	提供降级服务	非核心功能
限流	限制请求数量	流量突增

熔断器实现

class CircuitBreaker { constructor(options = {}) { this.failureThreshold = options.failureThreshold || 5; this.successThreshold = options.successThreshold || 3; this.timeout = options.timeout || 30000; this.state = 'closed'; this.failureCount = 0; this.successCount = 0; this.lastFailureTime = null; } async execute(fn) { if (this.state === 'open') { if (Date.now() - this.lastFailureTime > this.timeout) { this.state = 'half-open'; } else { throw new Error('Circuit breaker is open'); } } try { const result = await fn(); this.onSuccess(); return result; } catch (error) { this.onFailure(); throw error; } } onSuccess() { if (this.state === 'half-open') { this.successCount++; if (this.successCount >= this.successThreshold) { this.state = 'closed'; this.successCount = 0; this.failureCount = 0; } } else { this.failureCount = 0; } } onFailure() { this.failureCount++; this.lastFailureTime = Date.now(); if (this.failureCount >= this.failureThreshold) { this.state = 'open'; } } } // 使用 const breaker = new CircuitBreaker(); async function fetchData() { return breaker.execute(async () => { const response = await fetch('/api/data'); return response.json(); }); }

负载均衡

负载均衡算法

class LoadBalancer { constructor(servers) { this.servers = servers; this.index = 0; this.weights = servers.map(() => 1); } // 轮询 roundRobin() { const server = this.servers[this.index]; this.index = (this.index + 1) % this.servers.length; return server; } // 加权轮询 weightedRoundRobin() { let totalWeight = this.weights.reduce((a, b) => a + b, 0); let random = Math.random() * totalWeight; for (let i = 0; i < this.servers.length; i++) { random -= this.weights[i]; if (random <= 0) { return this.servers[i]; } } return this.servers[0]; } // 最少连接 leastConnections(connections) { let minConnections = Infinity; let selectedServer = this.servers[0]; for (const server of this.servers) { if (connections[server] < minConnections) { minConnections = connections[server]; selectedServer = server; } } return selectedServer; } }

数据一致性

CAP定理

CAP定理：分布式系统只能同时满足以下三点中的两点 C (Consistency) → 一致性 A (Availability) → 可用性 P (Partition tolerance) → 分区容错性

最终一致性

// 使用事件溯源实现最终一致性 class EventStore { constructor() { this.events = []; this.subscribers = []; } append(event) { this.events.push(event); for (const subscriber of this.subscribers) { subscriber(event); } } subscribe(callback) { this.subscribers.push(callback); } } // 使用 const eventStore = new EventStore(); eventStore.subscribe((event) => { if (event.type === 'orderCreated') { updateInventory(event.data.orderId); } }); eventStore.append({ type: 'orderCreated', data: { orderId: 'abc123' } });

实战案例：高可用架构

class HighAvailabilitySystem { constructor() { this.primary = null; this.secondary = null; this.loadBalancer = new LoadBalancer([]); this.circuitBreaker = new CircuitBreaker(); } async init() { this.primary = await this.startService('primary'); this.secondary = await this.startService('secondary'); this.loadBalancer = new LoadBalancer([this.primary, this.secondary]); this.setupHealthCheck(); this.setupFailover(); } setupHealthCheck() { setInterval(async () => { const primaryHealth = await this.checkHealth(this.primary); if (!primaryHealth) { await this.failover(); } }, 10000); } async failover() { console.log('Failing over to secondary...'); [this.primary, this.secondary] = [this.secondary, this.primary]; this.loadBalancer = new LoadBalancer([this.primary, this.secondary]); } async handleRequest(req) { return this.circuitBreaker.execute(async () => { const server = this.loadBalancer.roundRobin(); return server.handle(req); }); } }

最佳实践

1. 多数据中心部署

// 跨区域部署 const regions = ['us-west', 'us-east', 'eu-west']; function getServerByRegion(region) { return servers[region]; }

2. 自动化故障恢复

// 自动重启失败的服务 const service = new Service(); service.on('crash', async () => { console.log('Service crashed, restarting...'); await service.restart(); });