Atlas Flex Limits

You can deploy and Flex clusters only in a subset of regions on AWS, Google Cloud, and Azure. To learn more about supported cloud provider regions for Flex clusters, see:

• Amazon Web Services (AWS)

• Google Cloud Platform (GCP)

• Microsoft Azure

Atlas uses MongoDB 8.0 for Flex clusters.

You can't upgrade the MongoDB version that Flex clusters run. Atlas upgrades Flex clusters to the newest MongoDB version after several patch versions become available for that version. To learn more, see MongoDB Versioning.

You can't configure memory for Flex clusters.

You can't configure storage size for Flex clusters.

Flex clusters don't have pre-defined replica set tags.

You can't deploy a Flex cluster as a Sharded Cluster.

You can't perform primary failover testing on Flex clusters.

You can't perform regional outage testing on Flex clusters.

You can't configure database auditing on Flex clusters.

You can't configure encryption at Rest using Customer Key Management Flex clusters.

You can't configure network peering connections on Flex clusters.

You can't view the database access history for Flex clusters.

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

On Flex clusters, aggregation pipelines don't support the $currentOp, $listLocalSessions, $listSessions, or $planCacheStats stages.

Flex clusters aggregation pipelines can have a maximum of 50 stages.

Flex clusters can only trigger alerts configured with one of the following alert conditions:

• Connections

• Logical Size

• Network

• Opscounter

The Metrics view of a Flex cluster displays only the following metrics:

• Connections

• Logical Size

• Network

• Opscounter

To learn more, see Real Time Metrics.

Flex clusters support the following authentication methods only:

• Password (SCRAM-SHA1)

• X.509 Certificates

• AWS IAM

Flex clusters don't provide automatically scaling storage.

Flex clusters can store documents with a maximum of 50 nested levels.

Flex clusters don't support building indexes with a rolling build.

For Flex clusters, you can use only strings and regular expressions in filters on database names (namespace ns fields) in change streams. You can't use commands, such as $in, in database namespace filters. This limit doesn't apply to filtering on collection names in change streams.

For Flex clusters, you can't run mongorestore or mongodump on the admin database. If you use the --db option to set the destination database to admin, the program returns an error.

Flex clusters have a maximum of 500 connections.

Flex clusters can't use the noTimeout cursor option.

Changes to custom roles might take up to 30 seconds to deploy in Flex clusters.

Flex clusters have a maximum of 100 databases and a total of 500 collections.

Flex clusters don't allow:

• Read access to any collection in the local database with the exception of read access to the oplog.

  You must precede a $regex query on an oplog in an M0 Free cluster or M2/M5 Shared cluster with a caret (^). Otherwise, the following error occurs:

  MongoServerError: Oplog ns RegEx queries must begin with ^

• Write access to any collection in the local and config databases.

• Read or write access to any collection in the admin database.

Atlas issues an error similar to the following if you attempt to read or write to collections in these databases:

command <cmd name> is not allowed in this Atlas tier
(Unauthorized) not authorized on <db name> to execute command <cmd name>

Flex clusters don't allow you to download logs.

• Custom policies are not supported for M2 and M5 cluster snapshots. Atlas always takes a single daily snapshot at the same time, starting 24 hours after the cluster was created.

  If you require finer-grained backups, consider upgrading to an M10 or larger cluster tier.

• On-demand snapshots are not supported for M2 and M5 clusters.

• You can't restore M2 and M5 snapshots to a sharded cluster. You can only restore M2 and M5 snapshots to replica sets.

• You can't restore Serverless instance snapshots to M2 and M5 clusters.

• Starting with MongoDB 5.0, you can restore snapshots of clusters that run only the two most recent major versions of MongoDB to M2 and M5 clusters.

  For example:

  • You can restore snapshots taken from clusters that run MongoDB 4.4 to an M2 or M5 cluster that runs MongoDB 5.0.

  • You can't restore snapshots taken from clusters that run MongoDB versions earlier than 4.4 to an M2 or M5 cluster that runs MongoDB 5.0.

Flex clusters limit the maximum total data storage space to 5 GB.

Data storage space includes the number of bytes of all uncompressed BSON documents stored in all collections, plus the bytes stored in their associated indexes.

Flex clusters don't support server-side JavaScript. For example, $where and map-reduce are unsupported.

Flex cluster namespaces are limited to 95 bytes. Database names are limited to 38 bytes.

Flex clusters don't provide access to the Performance Advisor.

The percentage of time that a query is running over any five minute period must remain under 100% on Flex clusters.

Flex clusters don't provide access to the Real-Time Performance Panel.

Flex clusters limits sort in-memory to 32 MB.

Flex clusters limit the number of read and write operations to 500 operations per second.

Atlas handles clusters that exceed the operations per second rate limit as follows:

• Atlas triggers a sub-second cooldown period before resuming the cluster's operations on a given connection. If the queue is greater than the operations per second limit, operations might wait for more than a second in the queue.

• If the number of operations per second drops below the rate limit threshold, Atlas resumes processing of the queued operations on each connection before processing any new operations on that connection.