Compatibility Changes in MongoDB 6.0

allowDiskUse Changes

Starting in MongoDB 6.0, pipeline stages that require more than 100 megabytes of memory to execute write temporary files to disk by default. These temporary files last for the duration of the pipeline execution and can influence storage space on your instance. In earlier versions of MongoDB, you must pass { allowDiskUse: true } to individual find and aggregate commands to enable this behavior.

Individual find and aggregate commands can override the allowDiskUseByDefault parameter by either:

• Using { allowDiskUse: true } to allow writing temporary files out to disk when allowDiskUseByDefault is set to false

• Using { allowDiskUse: false } to prohibit writing temporary files out to disk when allowDiskUseByDefault is set to true

$$SEARCH_META Limitations

Starting in MongoDB 6.0, the Atlas Search $$SEARCH_META aggregation variable can be used anywhere after a $search stage in any pipeline, but it cannot be used after the $lookup or $unionWith stage in any pipeline. The $$SEARCH_META aggregation variable cannot be used in any subsequent stage after a $searchMeta stage.

Orphan Documents

Starting in MongoDB 5.3, during range migration, change stream events are not generated for updates to orphaned documents.

Resume Tokens

Starting in MongoDB 6.0.9, resume tokens created by change stream pipelines with the new $changeStreamSplitLargeEvent stage are incompatible with MongoDB 5.0. For details about resume tokens, see Resume Tokens.

Filters

Starting in MongoDB 6.0, whenever possible, match filters are applied to change streams earlier than in prior releases. This improves performance. However, when a filter is narrowly defined, an earlier match may cause an operation that succeeds in prior versions to fail in 6.0.

Last Remaining Shard Key Index Cannot be Dropped Inadvertently

Starting in MongoDB 6.0, passing "*" to dropIndexes or db.collection.dropIndexes() drops all indexes except for the _id index and the last remaining shard key index, if one exists. Attempts to explicitly drop the last remaining shard key index raise an error.

Existing Indexes Can Be Dropped During an Index Build

Starting in MongoDB 5.2, you can use dropIndexes or db.collection.dropIndexes() to drop existing indexes on the same collection even if there is an index build in progress. In earlier versions, attempting to drop a different index during an in-progress index build results in a BackgroundOperationInProgressForNamespace error.

2dsphere Document Index Keys

To prevent out of memory errors, indexMaxNumGeneratedKeysPerDocument limits the maximum number of 2dsphere index keys generated for a single document.

See indexMaxNumGeneratedKeysPerDocument.

Index Key Format

Starting in MongoDB 6.0, a change to the unique index key format was introduced. If you create a unique index in MongoDB 6.0, the index won't work with MongoDB versions earlier than 5.3.2 or 5.0.7.

Community Edition

• RHEL-72-s390x

$regex Find Queries No Longer Ignore Invalid Regex

Starting in MongoDB 5.1, invalid $regex options options are no longer ignored. This change makes $regex options more consistent with the use of $regex in the aggregate command and projection queries.

$regex Schema Validation Error Behavior

Starting in MongoDB 5.1, if a collection has schema validation rules that contain invalid $regex options the server:

• Prevents all insert and update operations until the schema validation rules containing the invalid regex pattern are modified with the collMod command.

• Writes a warning error to the mongod log file.

Assert Cluster Wide Write Concern is Set When Starting or Adding Shard

Starting in MongoDB 5.1, when starting, restarting or adding a shard server with sh.addShard() the Cluster Wide Write Concern (CWWC) must be set.

If the CWWC is not set and the shard is configured such that the default write concern is { w : 1 } the shard server will fail to start or be added and returns an error.

See default write concern calculations for details on how the default write concern is calculated.

rs.reconfig Cluster Wide Write Concern Validation

Starting in MongoDB 5.1, you must set the Cluster Wide Write Concern (CWWC) prior to issuing any reconfigs that would otherwise change the default write concern of the new replica set member.

Intra-Cluster Authentication

Starting in MongoDB 5.3, SCRAM-SHA-1 cannot be used for intra-cluster authentication. Only SCRAM-SHA-256 is supported.

In previous MongoDB versions, SCRAM-SHA-1 and SCRAM-SHA-256 can both be used for intra-cluster authentication, even if SCRAM is not explicitly enabled.

FIPS Mode Defaults SCRAM-SHA-1 Authentication to Off

Starting in MongoDB 5.1, instances running in FIPS mode have the SCRAM-SHA-1 authentication mechanism disabled by default. You can enable the SCRAM-SHA-1 authentication mechanism with the setParameter.authenticationMechanisms command.

This change will not affect drivers which target MongoDB setFeatureCompatibilityVersion 4.0+.

OCSP Must be Enabled

Starting in MongoDB 6.0, if ocspEnabled is set to true during initial sync, all nodes must be able to reach the OCSP responder.

If a member fails in the STARTUP2 state, set tlsOCSPVerifyTimeoutSecs to a value that is less than 5.

Secondary Indexes on Time Series Collections

If there are secondary indexes on time series collections and you need to downgrade the feature compatibility version (fCV), you must first drop any secondary indexes that are incompatible with the downgraded fCV. See setFeatureCompatibilityVersion.

Deprecations

$mod Error Behavior

Starting in MongoDB 5.1 (and 5.0.4), the $mod operator returns an error if the divisor or remainder values evaluate to certain values. See $mod behavior.

Legacy Opcodes Removed

MongoDB 6.0 removes support for the following legacy opcodes and database commands:

• OP_INSERT

• OP_DELETE

• OP_UPDATE

• OP_KILL_CURSORS

• OP_GET_MORE

• OP_QUERY

• getLastError

If you attempt to connect to a MongoDB 3.4 or older mongod instance with a MongoDB 5.1 or newer mongo shell, you will receive an error message like the following:

Connection handshake failed. Is your mongod 3.4 or older?
:: caused by :: network error while attempting to run command
'isMaster' on host '127.0.0.1:27017'

Removed Deprecated Array and String Functions for Server-Side JavaScript

MongoDB 6.0 upgrades the internal JavaScript engine used for server-side JavaScript, $accumulator, $function, and $where expressions and from MozJS-60 to MozJS-91. Several deprecated, non-standard array and string functions that existed in MozJS-60 are removed in MozJS-91.

For the complete list of removed array and string functions, see the next sections on this page.

Default db.stats() Settings

Starting in MongoDB 6.0, the dbStats command and the db.stats() method only report free space assigned to collections if the freeStorage parameter is set to 1.

Index Filters and Collations

Starting in MongoDB 6.0, an index filter uses the collation previously set using the planCacheSetFilter command.

Arrays in Collections and Views with distinct Command

Starting in MongoDB 6.0, the distinct command returns the same results for collections and views when using arrays.

See Arrays in Collections and Views.

Clustered Collections

Starting in MongoDB 5.3, if you are using clustered collections, you must drop those collections before you can downgrade to an earlier MongoDB version.

User Write Blocking

Starting in MongoDB 6.0, if you need to downgrade the feature compatibility version, ensure you disable cluster-to-cluster replication and user write blocking.

See Cluster-to-Cluster Sync and User Write Blocking.

Time Series Collections

You must drop time series collections before downgrading:

• MongoDB 6.0 or later to MongoDB 5.0.7 or earlier.

• MongoDB 5.3 to MongoDB 5.0.5 or earlier.

See Time Series Collections.

Cluster Parameters

Starting in MongoDB 6.0, ensure that all setClusterParameter operations have completed. fCV downgrade cannot occur successfully if there are any ongoing setClusterParameter operations on sharded clusters.

SELinux Policy Data

Starting in MongoDB 5.1, you must run the following command from the directory into which the SELinux policy was previously cloned before you can downgrade to an earlier MongoDB version:

sudo make uninstall

See:

• Configure SELinux Policy Data Directory

• Configure Enterprise SELinux Policy Data Directory

Key Management Interoperability Protocol (KMIP) Settings

Starting in MongoDB 6.0, the default KMIP protocol version is 1.2. To use KMIP version 1.0 or 1.1, use the useLegacyProtocol setting.

Starting in MongoDB 5.3 Enterprise, if you are using the following KMIP settings, you must remove them from the configuration file before you can downgrade to an earlier MongoDB version:

• security.kmip.keyStatePollingSeconds

• security.kmip.activateKeys

Time-based Retention of Change Streams Pre- and Post-Image Collections

Starting in MongoDB 6.0, if you are using changeStreamOptions.preAndPostImages.expireAfterSeconds to control time-based retention of change streams pre- and post-image collections, you must ensure there are no active setClusterParameter operations when downgrading.

Audit Log Encryption Settings

Starting in MongoDB 6.0 Enterprise, if you are using audit log encryption, you must remove the following settings from the configuration file before you can downgrade to an earlier MongoDB version:

• auditLog.auditEncryptionKeyIdentifier

• auditLog.localAuditKeyFile

Existing encrypted audit logs remain encrypted, and you can keep any procedures you have developed for storage and processing of encrypted logs.

See Audit Log.

Change Streams with Document Pre- and Post-Images

Starting in MongoDB 6.0, if you are using document pre- and post-images for change streams, you must disable changeStreamPreAndPostImages for each collection using the collMod command before you can downgrade to an earlier MongoDB version.

Change Streams with Expanded Events

If your application uses change streams, ensure that it does not require the showExpandedEvents option, which will not be available after downgrade.

LDAP with srv: and srv_raw:

If your cluster's configuration is using the new "srv:" or "srv_raw:" URL types in its LDAP configuration, it will be unable to restart after a downgrade. Remove the new URL types from your cluster's configuration before downgrading.

Collections with Encrypted Fields

You must drop collections that use encrypted fields before you can complete the fCV downgrade. The downgrade will not complete if there are collections using encryptedFields.