Clusters, alerts & events
Cluster configuration, backup snapshots and search indexes — plus the project’s alert list and activity feed.
Clusters
The Overview leads with a green/amber/red health hero, computed from open unacknowledged alerts and any cluster that is not idle, then a project summary and a state tile per cluster.
Tapping a cluster opens its page:
- its configuration summary — tier, region, version, topology;
- cloud-backup snapshots, the ten most recent;
- Atlas Search and Vector Search indexes.
The last two are wrapped so that not enabled reads differently from forbidden: a 404 means the feature is not turned on, while a 401/403 says plainly that only Project Owners can view backups or Atlas Search. Those are different problems with different fixes, and collapsing them into “error” wastes your time.
A search index row opens a detail page with its metadata and the full index definition — mappings and fields — as a JSON tree.
Alerts
GET /alerts — Atlas’ own warnings, either open only or all, newest first, with
acknowledged ones muted rather than hidden. This is the list the health hero is built from,
so tapping through from the Overview lands exactly where you expect.
Events
GET /events — the project activity feed, newest first, with a free-text search. Who
changed the cluster tier, when the last backup ran, which key was rotated: the audit
questions that come up right after an incident, rather than during it.
Back to MongoDB Atlas.