Manual Restore One Snapshot


This feature is not available for M0 (Free Tier) clusters. For more information, see Atlas M0 (Free Tier) Limitations.

Atlas provides a mechanism for downloading cloud provider, legacy backup, and M2/M5 snapshots as compressed files.

The downloaded files consist of the raw files copied from the data directory. mongorestore is incompatible with these files. To access the data files, use the following procedure to start a mongod instance and point it to the extract directory.

To learn more about this backup type, see Cloud Backups.

When manually downloading a cloud backup snapshot, there is the possibility that your download may be interrupted for some reason. Atlas keeps the request alive and allows it to be restarted as long as the most recent download failure is no more than one hour old.

  1. Click Clusters.
  2. Click the name of your Atlas Cluster.
  3. Click the Backup tab.
  4. Click the Snapshots sub-tab.
  5. Click Download.

    Atlas generates a one-time use download link that expires within 1 hour after its creation.

    The amount of time to create this link increases with the size of the Atlas cluster.

Once the download is ready, Atlas:

  • Emails you an alert that your snapshot download is ready.
  • Displays the download link in the Restores & Downloads tab.
Available via API

As another option, you can request a restore snapshot using the API.


If the current project IP access list ranges do not cover the target client IP or CIDR address, click Add or Modify your IP Addresses to make changes to your Atlas project IP access list.

  1. Click Clusters.
  2. Click the name of your Atlas Cluster.
  3. Click the Backup tab.
  4. Click the Restores & Downloads sub-tab.
  5. Navigate to the restore snapshot you created.
  6. Click Download.

You can't download your snapshot if a prior snapshot download link hasn't been clicked.


Atlas compresses the snapshot into a .tar.gz file. This archive includes the snapshot and the mongod logs.

Once extracted, you can access the data files by starting a mongod instance on the host and pointing it at the extract directory using the --dbpath option.


The following command uses the tar utility to extract a tar archive with gzip compression:

tar -xvzf ~/Downloads/mongodb-snapshots/my-cluster-snapshot.tar.gz

The following command starts a mongod instance using the extracted data file directory:

mongod --dbpath ~/Downloads/mongodb-snapshots/my-cluster-snapshot/
Give Feedback