A snapshot is a stored point-in-time copy of database data used to preserve consistent recovery points. WiredTiger uses MVCC to provide a consistent view, then writes snapshot data into durable checkpoints across data files. section 8.1.1
Use snapshots with these operational considerations:
- MongoDB schedules checkpoints every 60 seconds; the previous checkpoint remains valid until metadata is atomically updated. section 8.1.1
- Increase minSnapshotHistoryWindowInSeconds when longer snapshot history is needed, but expect greater disk usage in high-volume environments. section 8.1.1
- Restore Atlas clusters from scheduled or on-demand snapshots, optionally targeting another cluster for testing or disaster recovery. section 21.1.2
- Expect Atlas restoration to clear the target, load the snapshot or replay oplog entries, take the cluster offline, and bring it back after verification. section 21.1.2
- Use mongodump and mongorestore for M0 clusters because free M0 clusters do not provide cloud backup snapshots. section 21.1.2
Restore a snapshot to another Atlas cluster
Run `atlas backups restores start automated --clusterName "MongoDB-in-Action-M10" --targetClusterName "MongoDB-in-Action-M10" --snapshotId 6725be13eae896510f35a553 --targetProjectId 65d70c5bc9b5633e80a9c998` to initiate an automated restore using a selected snapshot.
atlas backups restores start automated \ --clusterName "MongoDB-in-Action-M10" \ --targetClusterName "MongoDB-in-Action-M10" \ --snapshotId 6725be13eae896510f35a553 \ --targetProjectId 65d70c5bc9b5633e80a9c998
Browse a complete Atlas CLI command for restoring a snapshot to a target cluster. section 21.1.2
