Nestjs教程使用Prometheus进行监控与报警

在NestJS中使用Prometheus进行监控和报警的步骤如下：

1. 安装依赖
   首先需要安装[@willsoto](/user/willsoto)/nestjs-prometheus包来集成Prometheus。

   npm install [@willsoto](/user/willsoto)/nestjs-prometheus
   

2. 配置模块
   在你的NestJS项目中引入PrometheusModule并配置它。

   import { Module } from '[@nestjs](/user/nestjs)/common';
   import { PrometheusModule } from '[@willsoto](/user/willsoto)/nestjs-prometheus';
   
   [@Module](/user/Module)({
     imports: [
       PrometheusModule.register({
         defaultLabels: { app: 'my-app' },
       }),
     ],
   })
   export class AppModule {}
   

3. 创建指标
   使用装饰器定义监控指标。

   import { Controller, Get, MetricsController, promMetrics } from '[@willsoto](/user/willsoto)/nestjs-prometheus';
   import { Counter } from '[@willsoto](/user/willsoto)/nestjs-prometheus/build/src/prom.types';
   
   [@Controller](/user/Controller)()
   export class AppController {
     @Counter({ name: 'http_requests_total', help: 'Total number of HTTP requests' })
     private readonly httpRequests;
   
     constructor() {
       this.httpRequests = promMetrics().getCounter('http_requests_total');
     }
   
     @Get()
     root() {
       this.httpRequests.inc();
       return { message: 'Hello World!' };
     }
   }
   

4. 启动服务
   启动NestJS应用后，访问/metrics即可获取Prometheus格式的数据。

5. 设置报警规则
   在Prometheus配置文件中添加报警规则，例如：

   groups:
     - name: example
       rules:
         - alert: HighRequestRate
           expr: increase(http_requests_total[5m]) > 100
           for: 1m
           labels:
             severity: warning
           annotations:
             summary: "High request rate detected"
   

通过以上步骤，你可以实现NestJS应用的监控和报警功能。

要使用Prometheus监控NestJS应用，首先安装@nestjs/microservices和prom-client库。创建一个服务，用于暴露监控指标。

1. 安装依赖：

npm install @nestjs/microservices prom-client

1. 创建监控服务：

import { Injectable } from '@nestjs/common';
import { Counter, Gauge, Histogram, Registry } from 'prom-client';

@Injectable()
export class MetricsService {
  private readonly _requestCounter = new Counter({
    name: 'http_requests_total',
    help: 'Total HTTP requests',
    labelNames: ['method', 'endpoint', 'status_code'],
  });

  private readonly _responseTimeHistogram = Histogram.build({
    name: 'http_response_time_ms',
    help: 'HTTP response time in milliseconds',
    labelNames: ['method', 'endpoint'],
    buckets: [50, 100, 200, 300, 400, 500],
  });

  constructor() {
    // 注册到全局Registry
    Registry.setDefaultLabels({ app: 'my-nest-app' });
    Registry.register(this._requestCounter);
    Registry.register(this._responseTimeHistogram);
  }

  // 在请求处理时更新计数器
  trackRequest(method: string, endpoint: string, statusCode: number) {
    this._requestCounter.inc({ method, endpoint, status_code: statusCode.toString() });
  }

  // 在响应时记录耗时
  recordResponseTime(method: string, endpoint: string, durationMs: number) {
    this._responseTimeHistogram.observe({ method, endpoint }, durationMs);
  }
}

1. 在全局拦截器中使用该服务记录数据：

import { Injectable, NestInterceptor, ExecutionContext, CallHandler } from '@nestjs/common';
import { Observable } from 'rxjs';
import { finalize } from 'rxjs/operators';
import { MetricsService } from './metrics.service';

@Injectable()
export class MetricsInterceptor implements NestInterceptor {
  constructor(private readonly metricsService: MetricsService) {}

  intercept(context: ExecutionContext, next: CallHandler): Observable<any> {
    const request = context.switchToHttp().getRequest();
    const { method, url } = request;
    const start = Date.now();

    return next.handle().pipe(
      finalize(() => {
        const duration = Date.now() - start;
        const statusCode = request.res.statusCode;
        this.metricsService.trackRequest(method, url, statusCode);
        this.metricsService.recordResponseTime(method, url, duration);
      }),
    );
  }
}

1. 在AppModule中注册拦截器：

import { Module, MiddlewareConsumer, RequestMethod } from '@nestjs/common';
import { AppController } from './app.controller';
import { AppService } from './app.service';
import { MetricsInterceptor } from './metrics.interceptor';

@Module({
  imports: [],
  controllers: [AppController],
  providers: [AppService, MetricsService],
})
export class AppModule {
  configure(consumer: MiddlewareConsumer) {
    consumer
      .apply(MetricsInterceptor)
      .exclude({ path: '*', method: RequestMethod.GET })
      .forRoutes('*');
  }
}

1. 启动NestJS应用后，Prometheus可以通过抓取/metrics路径获取数据。配置Prometheus Job来抓取你的应用数据，并设置警报规则。