Better Stack Automated Maintenance During Deployments

[fe51]

Pitch

Automate Better Stack maintenance windows during Ansible deployments to:

• avoid false downtime alerts
• keep clean uptime stats
• keep deployment traceability
• remove manual maintenance actions

Deployment Flow

1. Start short maintenance window ( YY min -> to be defined, and might depend on the service)
2. Deploy new image/version
3. Wait for healthcheck (/health) ?
4. End maintenance early or let it expire automatically

---

Better Stack Targeting

Deployment must notify the correct Better Stack monitor depending on (as an example)

Service	Environment	Target
frontend	prod	prod frontend monitor
backend	prod	prod backend monitor
frontend	dev	dev frontend monitor
backend	dev	dev backend monitor

Use Ansible variables/inventory, no hardcoded IDs.

Example:

service_name: frontend
environment: prod
betterstack_monitor_id: xxx
betterstack_api_token: vault_secret

Specs suggestion

Failure Handling

If deploy fails:

• maintenance stays active until expiry
• playbook fails normally
• optional notification later

Security

API token stored in Ansible Vault / CI secrets

ansible ideas

• Reusable role: roles/betterstack_maintenance

• usage

pre_tasks:
  - include_role:
      name: betterstack_maintenance
    vars:
      state: start

roles:
  - deploy_app

post_tasks:
  - include_role:
      name: betterstack_maintenance
    vars:
      state: stop

• maintenance windows instead choice over pause/unpause because they suppress false alerts while preserving a clear audit trail and deployment history