Docs Menu

Serverless Instance Limitations

On this page

Serverless instances are in a preview release and do not currently support the Atlas features listed below. If you require these capabilities, please use a dedicated cluster.

MongoDB plans to add support for more configurations and capabilities on serverless instances over time. A checkmark below indicates that MongoDB plans to support the feature for serverless instances in the future.

Coming Soon
Multi-Region Deployments
Multi-Cloud Deployments
Sharded Deployments

Advanced Enterprise Security Features (including LDAP and Database Auditing)

Serverless instances do support X.509 certificates and IAM for authentication.

Coming Soon
Migrate into Clusters from Atlas Serverless instances
Live Migrate into Atlas Serverless instances

Configure Alerts on Service Metrics Billing Metrics

Atlas supports configuring alerts for your project or organization if your bill exceeds a certain threshold.

Configure Backups

While serverless instances do allow backups, you cannot change the backup policies.

Perform Point-in-Time or Automated Restores from Backup Snapshots
Use Predefined Replica Set Tags
Track Database Access
Use Server-Side JavaScript, such as $where, $function, $accumulator and map-reduce
Download Database Logs
Use Wire Compression Between Clients and Atlas Serverless instances

Additionally, serverless instances have the following operational limitations and considerations:


Serverless instances don't support the allowDiskUse option for the aggregation command, its helper method, or the cursor.allowDiskUse() query cursor method.

Serverless instances don't support the $out stage. Use $merge instead.

Aggregation fields on serverless instances that represent database and collection names (such as $merge values) can't be expressions.

Aggregation pipelines for serverless instances can have a maximum of 50 stages.

The $sort stage has a limit of 32 megabytes of RAM.
Serverless instances don't routinely cap operation throughput. Atlas may throttle operations for your serverless instance temporarily while the system scales.
Serverless instances can support up to 500 simultaneous connections.
Database Commands

Some database commands have limitations for serverless instances. To learn more, see Unsupported Commands in Serverless Instances.

You cannot create a capped collection or convert an existing collection to a capped collection.

Namespaces and Database Names
Atlas limits serverless instance namespaces to 95 characters and database names to 38 characters.
Database and Collections
Serverless instances have a maximum of 50 databases and 500 collections total.
Custom Roles
Changes to custom roles may take up to 30 seconds to deploy in serverless instances.
Reading from the Oplog
Serverless instances don't provide direct read access to the oplog or any other collection in the local database.
Serverless instances don't support change streams.
Serverless instances don't support collation on collections, indexes, or queries.
BSON Nested Object Depth
Serverless instances can't store documents with more than 50 nested levels.
Transaction Size
Serverless instances support multi-document transactions that are up to 700 MB in size. Atlas aborts any serverless instance transactions that exceed 700 MB.

To connect to your serverless instance using a driver, you must use at least one of the following versions:

1.11.0 (Extension)
1.10.0 (Library)

To connect to serverless instances using mongosh, you must use version 1.0.0 or later.


You can't connect to serverless instances using the legacy mongo shell.

To connect to serverless instances using MongoDB Compass, you must use version 1.28 or later.

To import data using the MongoDB Tools, including mongodump, mongorestore, mongoexport, and mongoimport, you must have MongoDB Tools version 100.5.x or later.

←  Atlas LimitationsAtlas M0 (Free Cluster), M2, and M5 Limitations →
Give Feedback
© 2022 MongoDB, Inc.


  • Careers
  • Investor Relations
  • Legal Notices
  • Privacy Notices
  • Security Information
  • Trust Center
© 2022 MongoDB, Inc.