Alert Suppression Limits in MongoDB Atlas

Goal: Reduce Notification Noise

In a production environment, the primary objective is to trigger actionable alerts while avoiding repeated or irrelevant notifications that can lead to alert fatigue. This requires a clear understanding of how MongoDB Atlas’s suppression mechanisms work and where they fall short.

Current Constraints

• Suppression cooldown applies only to the same alert rule; simultaneous breaches of distinct rules still generate notifications.
• No per‑metric suppression for anomaly‑based alerts.
• There is no automated silencing during scheduled maintenance windows.
• Programmatic creation of alerts via a public REST API is not yet available.

Unresolved Questions

1. Does the suppression cooldown affect all alerts that share the same metric but belong to different rules, or is it strictly per‑rule?
2. Can a maintenance window be defined that globally silences all alerts across the cluster, and if not, what manual steps are required?
3. Is there an upcoming feature or Terraform provider support that will expose programmatic alert creation and granular per‑metric suppression?