server/src/services/PerformanceMonitorService.ts

import Logger from '../utils/Logger'

/**
 * 性能指标接口
 */
export interface PerformanceMetrics {
  timestamp: number
  memoryUsage: MemoryMetrics
  connectionMetrics: ConnectionMetrics
  messageMetrics: MessageMetrics
  systemMetrics: SystemMetrics
}

/**
 * 内存指标
 */
export interface MemoryMetrics {
  heapUsed: number      // MB
  heapTotal: number     // MB
  external: number      // MB
  rss: number          // MB
  heapUsedPercent: number
}

/**
 * 连接指标
 */
export interface ConnectionMetrics {
  totalConnections: number
  activeConnections: number
  idleConnections: number
  newConnectionsPerMinute: number
  disconnectionsPerMinute: number
}

/**
 * 消息指标
 */
export interface MessageMetrics {
  messagesPerSecond: number
  averageLatency: number  // ms
  p95Latency: number      // ms
  p99Latency: number      // ms
  errorRate: number       // %
}

/**
 * 系统指标
 */
export interface SystemMetrics {
  uptime: number          // seconds
  cpuUsage: number        // %
  eventLoopLag: number    // ms
}

/**
 * 性能监控服务
 */
export class PerformanceMonitorService {
  private logger = new Logger('PerformanceMonitor')
  
  // 指标收集
  private metrics: PerformanceMetrics[] = []
  private readonly MAX_METRICS_HISTORY = 60  // 保留最近60条记录
  
  // 消息延迟追踪
  private messageLatencies: number[] = []
  private readonly MAX_LATENCY_SAMPLES = 1000
  
  // 连接统计
  private connectionsPerMinute = 0
  private disconnectionsPerMinute = 0
  private lastConnectionCount = 0
  
  // 消息统计
  private messagesThisSecond = 0
  private messagesLastSecond = 0
  private errorsThisSecond = 0
  private errorsLastSecond = 0
  
  // 事件循环监控
  private lastEventLoopCheck = Date.now()
  private eventLoopLag = 0

  constructor() {
    this.startMonitoring()
  }

  /**
   * 记录消息延迟
   */
  recordMessageLatency(latency: number): void {
    this.messageLatencies.push(latency)
    if (this.messageLatencies.length > this.MAX_LATENCY_SAMPLES) {
      this.messageLatencies.shift()
    }
  }

  /**
   * 记录消息
   */
  recordMessage(): void {
    this.messagesThisSecond++
  }

  /**
   * 记录错误
   */
  recordError(): void {
    this.errorsThisSecond++
  }

  /**
   * 记录连接
   */
  recordConnection(): void {
    this.connectionsPerMinute++
  }

  /**
   * 记录断开连接
   */
  recordDisconnection(): void {
    this.disconnectionsPerMinute++
  }

  /**
   * 获取当前性能指标
   */
  getCurrentMetrics(): PerformanceMetrics {
    const memUsage = process.memoryUsage()
    const heapUsedMB = Math.round(memUsage.heapUsed / 1024 / 1024)
    const heapTotalMB = Math.round(memUsage.heapTotal / 1024 / 1024)
    const externalMB = Math.round(memUsage.external / 1024 / 1024)
    const rssMB = Math.round(memUsage.rss / 1024 / 1024)

    const metrics: PerformanceMetrics = {
      timestamp: Date.now(),
      memoryUsage: {
        heapUsed: heapUsedMB,
        heapTotal: heapTotalMB,
        external: externalMB,
        rss: rssMB,
        heapUsedPercent: Math.round((heapUsedMB / heapTotalMB) * 100)
      },
      connectionMetrics: {
        totalConnections: 0,  // 由调用者设置
        activeConnections: 0,
        idleConnections: 0,
        newConnectionsPerMinute: this.connectionsPerMinute,
        disconnectionsPerMinute: this.disconnectionsPerMinute
      },
      messageMetrics: {
        messagesPerSecond: this.messagesLastSecond,
        averageLatency: this.calculateAverageLatency(),
        p95Latency: this.calculatePercentileLatency(95),
        p99Latency: this.calculatePercentileLatency(99),
        errorRate: this.messagesLastSecond > 0
          ? Math.round((this.errorsLastSecond / this.messagesLastSecond) * 100 * 100) / 100
          : 0
      },
      systemMetrics: {
        uptime: Math.round(process.uptime()),
        cpuUsage: this.calculateCpuUsage(),
        eventLoopLag: this.eventLoopLag
      }
    }

    return metrics
  }

  /**
   * 计算平均延迟
   */
  private calculateAverageLatency(): number {
    if (this.messageLatencies.length === 0) return 0
    const sum = this.messageLatencies.reduce((a, b) => a + b, 0)
    return Math.round(sum / this.messageLatencies.length * 100) / 100
  }

  /**
   * 计算百分位延迟
   */
  private calculatePercentileLatency(percentile: number): number {
    if (this.messageLatencies.length === 0) return 0
    const sorted = [...this.messageLatencies].sort((a, b) => a - b)
    const index = Math.ceil((percentile / 100) * sorted.length) - 1
    return sorted[Math.max(0, index)]
  }

  /**
   * 计算CPU使用率 (简化版)
   */
  private calculateCpuUsage(): number {
    // 这是一个简化的实现，实际应该使用 os.cpus() 或专门的库
    const usage = process.cpuUsage()
    return Math.round((usage.user + usage.system) / 1000000 * 100) / 100
  }

  /**
   * 启动监控任务
   */
  private startMonitoring(): void {
    // 每秒更新消息统计
    setInterval(() => {
      this.messagesLastSecond = this.messagesThisSecond
      this.errorsLastSecond = this.errorsThisSecond
      this.messagesThisSecond = 0
      this.errorsThisSecond = 0
    }, 1000)

    // 每分钟重置连接统计
    setInterval(() => {
      this.connectionsPerMinute = 0
      this.disconnectionsPerMinute = 0
    }, 60000)

    // 每10秒收集一次完整指标
    setInterval(() => {
      const metrics = this.getCurrentMetrics()
      this.metrics.push(metrics)
      
      if (this.metrics.length > this.MAX_METRICS_HISTORY) {
        this.metrics.shift()
      }

      this.logMetrics(metrics)
    }, 10000)

    // 监控事件循环延迟
    this.monitorEventLoopLag()
  }

  /**
   * 监控事件循环延迟
   */
  private monitorEventLoopLag(): void {
    let lastCheck = Date.now()
    
    setInterval(() => {
      const now = Date.now()
      const expectedDelay = 1000  // 1秒
      const actualDelay = now - lastCheck
      this.eventLoopLag = Math.max(0, actualDelay - expectedDelay)
      lastCheck = now
    }, 1000)
  }

  /**
   * 输出指标日志
   */
  private logMetrics(metrics: PerformanceMetrics): void {
    const mem = metrics.memoryUsage
    const msg = metrics.messageMetrics
    const conn = metrics.connectionMetrics
    const sys = metrics.systemMetrics

    this.logger.info(`
📊 性能指标 (${new Date(metrics.timestamp).toLocaleTimeString()}):
  💾 内存: ${mem.heapUsed}MB / ${mem.heapTotal}MB (${mem.heapUsedPercent}%) | RSS: ${mem.rss}MB
  📨 消息: ${msg.messagesPerSecond}/s | 延迟: ${msg.averageLatency}ms (p95: ${msg.p95Latency}ms, p99: ${msg.p99Latency}ms) | 错误率: ${msg.errorRate}%
  🔌 连接: ${conn.totalConnections}个 (活跃: ${conn.activeConnections}, 空闲: ${conn.idleConnections}) | 新增: ${conn.newConnectionsPerMinute}/min
  ⚙️  系统: 运行时间 ${sys.uptime}s | CPU: ${sys.cpuUsage}% | 事件循环延迟: ${sys.eventLoopLag}ms
    `)
  }

  /**
   * 获取历史指标
   */
  getMetricsHistory(limit: number = 10): PerformanceMetrics[] {
    return this.metrics.slice(-limit)
  }

  /**
   * 获取性能警告
   */
  getPerformanceWarnings(): string[] {
    const warnings: string[] = []
    const latest = this.metrics[this.metrics.length - 1]

    if (!latest) return warnings

    // 内存警告
    if (latest.memoryUsage.heapUsedPercent > 80) {
      warnings.push(`⚠️ 内存使用过高: ${latest.memoryUsage.heapUsedPercent}%`)
    }

    // 延迟警告
    if (latest.messageMetrics.p99Latency > 500) {
      warnings.push(`⚠️ 消息延迟过高: P99=${latest.messageMetrics.p99Latency}ms`)
    }

    // 错误率警告
    if (latest.messageMetrics.errorRate > 5) {
      warnings.push(`⚠️ 错误率过高: ${latest.messageMetrics.errorRate}%`)
    }

    // 事件循环延迟警告
    if (latest.systemMetrics.eventLoopLag > 100) {
      warnings.push(`⚠️ 事件循环延迟过高: ${latest.systemMetrics.eventLoopLag}ms`)
    }

    return warnings
  }

  /**
   * 获取性能报告
   */
  getPerformanceReport(): string {
    const warnings = this.getPerformanceWarnings()
    const latest = this.metrics[this.metrics.length - 1]

    if (!latest) return '暂无数据'

    let report = '📈 性能报告\n'
    report += '='.repeat(50) + '\n'
    report += `时间: ${new Date(latest.timestamp).toLocaleString()}\n`
    report += `内存: ${latest.memoryUsage.heapUsed}MB / ${latest.memoryUsage.heapTotal}MB\n`
    report += `消息吞吐: ${latest.messageMetrics.messagesPerSecond}/s\n`
    report += `平均延迟: ${latest.messageMetrics.averageLatency}ms\n`
    report += `连接数: ${latest.connectionMetrics.totalConnections}\n`
    report += `运行时间: ${latest.systemMetrics.uptime}s\n`

    if (warnings.length > 0) {
      report += '\n⚠️ 警告:\n'
      warnings.forEach(w => report += `  ${w}\n`)
    } else {
      report += '\n✅ 系统运行正常\n'
    }

    return report
  }

  /**
   * 清理资源
   */
  destroy(): void {
    this.metrics = []
    this.messageLatencies = []
  }
}