On this page
Atlas supports deploying clusters onto Amazon Web Services, the Google Cloud Platform, and Microsoft Azure. The choice of cloud service provider and region or regions for the Atlas project affects the cost of running a Atlas cluster.
Multi-region cluster costs depend on the number of and location of additional regions selected. When creating a cluster, Atlas displays the cluster tier cost based on only the Preferred Region of the cluster. You can see the total cost of running the cluster in the Cluster Overview.
To learn more about configuring your cloud provider and region, see Cloud Providers and Regions.
Atlas provides different cluster tiers. Each cluster tier has a default RAM capacity, storage capacity, and maximum storage speed. The cluster's per-hour charge includes these default values. Atlas uses the selected cluster when deploying all the data-bearing  servers in your cluster.
Depending on the choice of cloud service provider, Atlas provides customization options for cluster storage capacity and the speed of that storage. If you add capacity or speed, you incur additional costs on top of the base cost. For multi-region clusters, the per-cluster cost, including any selected customizations, is relative to the Preferred Region. The Cluster Overview box shows your overall charges.
Atlas charges for storage capacity differently depending on whether you use the cluster default or specify a custom storage capacity.
If you use the default storage capacity, Atlas includes its cost in the cluster's per-hour cost.
If you customize the amount of storage capacity, Atlas charges for the full amount of storage. Atlas doesn't deduct the cost of the default storage capacity. This change can be a disk upgrade or a instance family change such as changing from general CPU to low-CPU.
M10 cluster defaults to 10 GB of storage. You can increase
this amount up to 120 GB of storage using this cluster tier.
If you increase the storage capacity to 50 GB, your monthly Atlas cost includes 50 GB of storage, not the cost of the additional 40 GB.
Increasing storage capacity can change the maximum IOPS available with each Custom Storage Speed.
Atlas measures storage speed as maximum IOPS. Each Atlas cluster tier offers a default storage speed that is included in the cluster's per-hour cost. The choice of cloud service provider and cluster affects the available storage speed customization options, as well as the cost of selecting a custom storage speed.
To help minimize cluster costs while retaining the flexibility to easily scale your cluster, you can enable Cluster Auto-Scaling. With auto-scaling, your cluster automatically scales its tier, storage capacity, or both in response to cluster usage. Auto-scaling reduces the need to manually optimize your cluster to adapt to your current workload.
Atlas charges for each replica set in a cluster.
For a replica set deployment, the backup costs are the cost to backup data for the replica set.
For a sharded cluster deployment, Atlas sums the cost to backup the data for each shard replica set and the config server replica set.
Atlas charges for the network transfer costs of restoring a snapshot. To learn how Atlas charges for network data, see Data Transfer.
The backup cost for legacy backups is calculated per GB per month. The monthly rate is annualized and then divided by 365 to arrive at a Daily Backup Rate per GB. The first 1 GB of backup data is free.
Monthly backup costs are primarily based on the size per-gigabyte of the data to back up. This size is roughly equivalent to the uncompressed size of all documents and all indexes for all the databases backed up.
To retrieve the size in gigabyte of the documents and indexes for a
given database, you can issue the
db.stats() method and sum the
db.stats(1024*1024*1024).dataSize + db.stats(1024*1024*1024).indexSize
The rate is based on your having 28 snapshots at steady state:
The six-hour snapshots are kept for two days;
The daily snapshots for one week,
The weekly snapshots for one month,
The monthly for one year.
Adding that up, you get 8 + 5 + 3 + 12 = 28 snapshots. We adjust the backup rate daily based on the following formula:
backupRatePerMonth = $1.25 + snapshotAtSteadyState/28 × $1.25
Changing the snapshot frequency or retention period affects the base rate per GB. See Snapshot Schedule for more information.
For a three-member replica set with 30GB of data, Atlas charges US$72.50 each month (using the default backup rate).
(30 GB - 1 GB) × US$2.50 = US$72.50
A sharded cluster with 3 shards contains 90GB of data, with each shard containing 30GB of data each. The config server replica set contains 5GB of data.
For this sharded cluster, Atlas charges US$227.50 per month (using the default backup rate).
((30 GB - 1 GB) × US$2.50) × 3) + ((5 GB - 1 GB) × US$2.50) = US$227.50
If you enabled Continuous Cloud Backups, Atlas bills you using the rates given for Continuous Cloud Backups.
Atlas takes 31 snapshots for Cloud Backup each month by default. You can change the backup policy if needed. In most cases, a new snapshot saves only the data that changed after your most recent snapshot.
Atlas takes one snapshot per shard. It takes the first snapshot in full and takes incremental snapshots afterward. Atlas can take more than one full snapshot per shard in certain cases. These cases include when:
Snapshot taken on a different node from the previous one due to failover.
Region priority changes for a multi-region cluster. This would set a different node as the one that creates snapshots. If Atlas deems the selected node as unhealthy, it tries to snapshot the next node that it favors.
Atlas can bill backup as high as the total storage capacity of the volume. This depends upon how the cloud provider stores the volume snapshots.
A storage volume on the cloud provider has a capacity of 4 TB. The cloud provider informs Atlas that the snapshots occupy the full volume capacity even though your backups only occupy 500 GB. Due to this reporting, Atlas bills you for 4 TB in backup storage.
If the existing snapshot storage volume becomes invalid, Atlas creates a new snapshot storage volume in the same region as the cluster's current backup node and takes a full-copy snapshot. Atlas continues using that backup node and its cloud provider region for snapshots and snapshot storage. This may result in a higher invoice for the few days required to re-establish incremental snapshots.
To learn more about how Atlas manages snapshot storage, see Back Up Your Database Deployment.
When restoring a cluster using a manual download via HTTPS, Atlas also charges for:
Each hour that the download link remains active (Atlas Backup Download VM charge), and
The total storage capacity of the restore virtual machine data volume (Atlas Backup Restore Storage charge).
You can't download Backup snapshots that you encrypted using KMS.
If you have questions on Cloud Backup backup sizing and pricing, please contact Atlas support by clicking Support from the left-hand navigation of the Atlas UI.
Cluster owners may enable Continuous Cloud Backup restores from Cloud Backups. PIT backups are billed based on the disk space occupied by an internal oplog, combined with the cloud backup snapshot size.
You can configure PIT backups to cover a window of time that you specify. Longer backup windows result in larger oplogs and higher backup costs.
Atlas takes 51 snapshots for Continuous Cloud Backup each month by default. You can change the backup policy if needed. In most cases, a new snapshot saves only the data that changed after your most recent snapshot.
Continuous Cloud Backups support incremental snapshots, where a new snapshot saves only the data that changed after your most recent snapshot. For example, a cluster with 10 GB of data and 3 snapshots may require less than 30 GB of total snapshot storage depending on how data changed between snapshots.
To compute the cost for Continuous Cloud Backups, Atlas obtains the raw metric data from the cloud providers and calculates the total size of all snapshots based on the region where it stores the snapshot and the amount of storage used per month:
If you enable Additional Backup Copies for your cluster, you incur charges for additional storage in other regions. For example, if you copy all of your backups to one additional region, you pay for approximately twice the storage: an amount for your standard backup, and approximately the same amount for your copies stored in the other region (storage costs vary by region).
Google Cloud Platform
In Google Cloud regions, the incrementality of your backups is not preserved. Each copy stored in Google Cloud is a full copy. Because of this, Google Cloud backup copies might incur higher costs than the original incremental backup copies.
In AWS and Azure regions, the incrementality of your backups is preserved during the copy process.
Lowering snapshot frequency or lowering snapshot retention lowers the cost per gigabyte. Increasing the snapshot frequency or the snapshot retention increases the cost per gigabyte. To modify the snapshot frequency or retention for a cluster, see Snapshot Schedule.
The cost of backups is dependent on the region of the replica set member targeted for snapshots. Modifying the region configuration for your cluster may reduce the cost per gigabyte for snapshot storage. To change regions, see scaling the cluster.
Excluding MongoDB Atlas Enterprise and MongoDB Atlas Platinum customers, if BI Connector for Atlas is enabled for your cluster:
The billing rate for the BI Connector for Atlas is described in the cluster console as a daily uplift on the cost of the associated cluster. You can view the rate when deploying your cluster or by modifying your cluster.
BI Connector for Atlas has a sustained-usage pricing. That is, the daily rate is charged only up to a maximum for the month.
Atlas charges the cluster cost and data storage cost for each data-bearing node  in your cluster.
For a replica set, the number of data-bearing nodes equals the replication factor.
For a sharded cluster, the number of data-bearing nodes equals the replication factor multiplied by the number of shards.
If you enable sharding, Atlas also runs three config servers in addition to your data-bearing nodes. Your selections for cluster tier and data storage do not affect the costs of the config nodes. Config servers are charged at a separate rate. Their cost is reflected in the cost of the cluster.
|||(1, 2) For replica sets, the data-bearing servers are the servers hosting the replica set nodes. For sharded clusters, the data-bearing servers are the servers hosting the shards. For sharded clusters, Atlas also deploys servers for the config servers; these are charged at a rate separate from the cluster costs.|