Datadog Monitoring

Guidelines for setting up application monitoring and observability.

stableGuideintermediatedatadogmonitoringobservability

1 min readUpdated Jul 24, 2026@observability-teamObservability

We use Datadog for system observability, logging, APM (Application Performance Monitoring), and alerting.

Agent Deployment

Datadog agents are run as a sidecar container in all ECS Fargate task definitions.
The agent automatically collects container metrics, resource utilization, and logs.

To enable APM tracing in Node.js applications, initialize the tracer as the very first line of your entry point file:

// dd-trace-init.js (Imported before any other modules)
const tracer = require('dd-trace').init({
  logInjection: true,
  analytics: true,
});

All application logs must be output to standard out (stdout) in JSON format.
Ensure log entries include context tags such as env, service, and version.
Avoid logging sensitive user data (PII) like passwords, email addresses, or authorization tokens.

All alert monitors must be defined as code using our Datadog Terraform provider. Alerts should route notifications to #ops-alerts on Slack.