Nokkel Monitoring Library

A monitoring library for integrating monitoring and observability platforms and generating logs for monitoring.

How to use the library

To install this library follow the steps below:

npm add @nokkel/monitoring

Once installed you can use the library to do the following:

logger.ts

import { createLogger } from '@nokkel/monitoring';

export const logger = createLogger({
  name: process.env.HOSTNAME,
  logLevel: process.env.LOG_LEVEL,

  // Can optionally configure the logger for the browser
  // isBrowser: true,
});

Enjoy!

Grafana Stack (Loki, Prometheus, Tempo)

Loki logging

logger.loki.ts

import { createLokiLogger } from '@nokkel/monitoring/grafana';

export const logger = createLokiLogger({
  name: process.env.SERVICE_NAME || 'my-service',
  logLevel: (process.env.LOG_LEVEL as any) || 'info',
  loki: {
    host: process.env.LOKI_URL || 'http://localhost:3100',
    labels: { service: 'my-service', env: process.env.NODE_ENV || 'dev' },
  },
});

logger.info({ msg: 'hello loki' });

Alternatively, compose your own pino with a Loki destination:

import { createLogger } from '@nokkel/monitoring';
import { createLokiDestination } from '@nokkel/monitoring/grafana';

const destination = createLokiDestination({ host: 'http://localhost:3100' });
export const logger = createLogger({ name: 'svc', logLevel: 'info', destination });

Prometheus metrics

Expose a /metrics endpoint Prometheus/Grafana Agent can scrape.

import { initMetrics, startMetricsServer, promClient } from '@nokkel/monitoring/metrics';

// set default labels and collect Node process metrics
initMetrics({ defaultLabels: { service: 'my-service' } });

// define custom metrics
export const requests = new promClient.Counter({ name: 'http_requests_total', help: 'Total HTTP requests' });

// start a metrics server on :9464/metrics
startMetricsServer({ port: 9464 });

OpenTelemetry tracing (Tempo)

Send traces via OTLP HTTP to Grafana Agent or Tempo.

import { initTracing } from '@nokkel/monitoring/tracing';

await initTracing({
  serviceName: 'my-service',
  serviceVersion: '1.0.0',
  environment: process.env.NODE_ENV || 'dev',
  otlpUrl: process.env.OTLP_TRACES_URL || 'http://localhost:4318/v1/traces',
});

// Your app code; auto-instrumentations cover http/express/etc.

Notes:

• Loki logging uses pino-loki. Ensure Grafana Loki accepts pushes at LOKI_URL.
• Metrics use prom-client and are scrape-based (no Pushgateway assumed).
• Traces use OpenTelemetry OTLP/HTTP; point OTLP_TRACES_URL to Grafana Agent/Tempo.