Skip to main content
Monitoring as Code (MaC) treats your monitoring setup like application code — versioned, reviewed, tested, and deployed through CI/CD. Instead of clicking through dashboards, you define monitors, alerts, and policies in configuration files.

​
The problem with dashboard-only monitoring

When monitoring configuration lives only in a web UI:
  • No version history — who changed that alert threshold last week?
  • No review process — changes go live without team review
  • No reproducibility — setting up a new environment means manual recreation
  • Configuration drift — staging and production diverge silently
  • Bus factor — only the person who clicked the buttons knows the setup

​
How Monitoring as Code works

MaC follows the same principles as Infrastructure as Code (IaC):
  1. Define monitors, channels, and policies in a declarative format (YAML, HCL, or code)
  2. Version the config in Git alongside your application code
  3. Review changes through pull requests with team review
  4. Test by validating syntax and previewing changes before applying
  5. Deploy automatically through CI/CD on merge to main

​
Benefits

​
Reproducibility

Spin up identical monitoring for a new environment in seconds: 
devhelm deploy -f devhelm.yml --yes
Every team member gets the same monitoring setup — no manual steps, no missed configurations.

​
Auditability

Git provides a complete history of every change: 
commit a1b2c3d — feat: add SSL certificate monitoring
commit d4e5f6a — fix: reduce API health check frequency to 30s
commit g7h8i9j — chore: add PagerDuty escalation for production

​
Review process

Monitoring changes go through the same PR review as code changes. Team members can catch issues like:
  • Missing alerts on critical services
  • Overly aggressive check frequencies
  • Incorrect assertion thresholds

​
Drift prevention

When someone changes a monitor through the dashboard, the next deploy reverts it to what’s in code. The YAML file is always the source of truth.

​
Disaster recovery

If you need to rebuild your monitoring from scratch, the config file has everything. No screenshots of dashboard settings, no tribal knowledge.

​
Declarative vs imperative

ApproachToolsHow it works
DeclarativeYAML config, Terraform HCLDescribe the desired state; the tool figures out what to create/update/delete
ImperativeSDK scripts, CLI commandsWrite step-by-step instructions to create each resource
Declarative is preferred for monitoring configuration because it handles drift, deletion, and idempotency automatically.

​
Getting started

DevHelm supports MaC through multiple tools:

YAML + CLI

Define everything in devhelm.yml — fastest path.

Terraform provider

Manage monitors alongside cloud infrastructure.

CI/CD patterns

Automate deploys on every push.

Migration guide

Move from dashboard-only to code-managed monitoring.