mongocli atlas backup restore start

The backup restore start command restores a deployment using the snapshot or point in time of a cluster. You can also restore a deployment using the Atlas UI or API.

mongocli atlas backup restore start <restore-job-type>
--clusterName <cluster-name>
[ --output|-o <output-format> ]
[ --oplogInc <oplog-Inc> ]
[ --oplogTs <oplog-timestamp> ]
[ --pointInTimeUTCMillis <pointInTime-UTC-ms> ]
[ --profile|-P <profile-name> ]
[ --projectId <project-ID> ]
--snapshotId <snapshot-ID>
[ --targetClusterName <cluster-name> ]
[ --targetProjectId <project-ID> ]

Use -h or --help to view the command-line help for this command.


If you create an automated or pointInTime restore job, Atlas removes all existing data on the target cluster prior to the restore.

Specify the type of restore job with one of the following arguments:

Atlas restores the snapshot corresponding to snapshotId to the Atlas cluster and project specified by targetClusterId and targetGroupId.

Atlas generates and displays a URL to download a .tar.gz of the snapshot corresponding to the snapshotId. The contents of the tar.gz archive contain the data files for your Atlas cluster.

See also:

To learn more about manually restoring the downloaded data files, see Manual Restore One Snapshot.

Download Limitations

Each cloud backup can have one download at a time, and each project can have a maximum of 20 downloads at a time.

The options required by or available for the command depend on the type of restore job:

The command prints the following fields in the specified format to the terminal if the command succeeds. If the command returns an error, see Troubleshooting for recommended solutions.

Give Feedback

On this page

  • Syntax
  • Arguments
  • Options
  • Output
  • Example