Datadog

The unified observability platform for metrics, traces, logs, and security. Datadog's Agent runs on your infrastructure, its APM auto-instruments your services, and its dashboards correlate everything into a single pane of glass — from host CPU to individual database queries.

What is Datadog?

Datadog is a SaaS observability platform that collects telemetry data — metrics, distributed traces, logs, events, and security signals — from every layer of your stack and surfaces it through unified dashboards, intelligent alerts, and AI-powered analysis. The Datadog Agent is a lightweight daemon you deploy on each host; it handles collection, enrichment, and forwarding.

Unlike open-source stacks where you compose Prometheus + Jaeger + ELK separately, Datadog provides a single agent, single backend, and single UI. The tradeoff is cost — Datadog is per-host SaaS, which becomes expensive at scale. The benefit is tight correlation: one click from a slow trace to the host metrics to the error logs from that exact request.

Key Capabilities

• Infrastructure Monitoring — CPU, memory, disk, network for hosts, containers, Kubernetes, cloud services
• APM & Distributed Tracing — end-to-end traces with automatic instrumentation for 20+ languages
• Log Management — ingestion, parsing, full-text search, log-to-trace correlation
• Synthetic Monitoring — browser tests and API checks that simulate user journeys
• Real User Monitoring (RUM) — actual user session data: page load, JS errors, user actions
• Database Monitoring — query-level performance for PostgreSQL, MySQL, MongoDB, Redis
• Cloud Cost Management — break down and optimize AWS/Azure/GCP spend
• Watchdog — ML anomaly detection that surfaces unusual patterns without threshold configuration

Agent Configuration

# /etc/datadog-agent/datadog.yaml (key settings)
api_key: 
site: datadoghq.com    # or datadoghq.eu for EU region

# Enable APM tracing
apm_config:
  enabled: true
  receiver_port: 8126

# Enable log collection
logs_enabled: true

# Enable process monitoring
process_config:
  process_collection:
    enabled: true

# Enable network performance monitoring
network_config:
  enabled: true

APM Auto-Instrumentation (Node.js)

// APM auto-instrumentation (Node.js)
// MUST be first import — before any other requires
import tracer from 'dd-trace';

tracer.init({
  service: 'payments-api',
  env: process.env.NODE_ENV,
  version: process.env.APP_VERSION,
  logInjection: true,         // correlate logs with traces
  runtimeMetrics: true,       // Node.js heap, GC, event loop metrics
  profiling: true,            // CPU/heap profiling
});

// Custom span
const span = tracer.startSpan('process-order', {
  childOf: tracer.scope().active(),
});
span.setTag('order.id', orderId);
span.setTag('order.value', amount);
try {
  await processOrder(orderId);
} catch (err) {
  span.setTag('error', err);
  throw err;
} finally {
  span.finish();
}

Agent Installation

# Ubuntu / Debian (one-liner)
DD_API_KEY= DD_SITE="datadoghq.com" \
  bash -c "$(curl -L https://install.datadoghq.com/scripts/install_script_agent7.sh)"

# Kubernetes (Helm)
helm repo add datadog https://helm.datadoghq.com
helm install datadog datadog/datadog \
  --set datadog.apiKey= \
  --set datadog.apm.portEnabled=true \
  --set datadog.logs.enabled=true

# Docker
docker run -d \
  -e DD_API_KEY= \
  -e DD_SITE=datadoghq.com \
  -e DD_APM_ENABLED=true \
  -v /var/run/docker.sock:/var/run/docker.sock:ro \
  -v /proc/:/host/proc/:ro \
  gcr.io/datadoghq/agent:7