Release Notes for MongoDB 8.2

8.2.7 - Upcoming

• SERVER-103990 Skip range deletions when transitioning from embedded to dedicated config servers

• SERVER-112273 Fix crash caused by killing an OplogWriter operation using killOp()

• SERVER-116284 Ensure MultiStatementTransactionRequestsSender sends commitTransaction to all shards before processing results

• SERVER-119413 Fix configTerm to handle int64 values when int32 is expected

• SERVER-119964 Collection catalog race condition shuts down oplog visibility thread after rollback

• WT-16375 Turn off failpoint when calling cursor->close()

• All JIRA issues closed in 8.2.7

• 8.2.7 Changelog

8.2.6 - Mar 17, 2026

• SERVER-106239: Investigate transaction participant state mismatch when running sharded aggregate with merging pipeline executing on a shard

• SERVER-110832: OplogWriter should not hold on to session through rollbacks

• SERVER-111007: Validation should distinguish between root causes when finding mixed schema data

• SERVER-113319: $match pushdown past a renamed field is done incorrectly when the expression root is a renamed field path expression

• SERVER-115962: Balancer does not make progress when the most loaded shard is already balanced within its zones

• SERVER-117623: Sharded multi-document transactions can observe partial effects of concurrent renameCollection operations

• SERVER-118428: mongocryptd rejects large messages

• SERVER-118849: Double free in SBE hash table spill

• SERVER-119317: Improve object lifecycle of MD5 hash state

• SERVER-119319: Fix ExpressionContext use-after-free bug

• WT-15739: Fix reconciliation leaking overflow pages

• All JIRA issues closed in 8.2.6

• 8.2.6 Changelog

8.2.5 - Feb 10, 2026

• SERVER-118711 - Can not create timeseries 2dsphere index with a non-system user

• All JIRA issues closed in 8.2.5

• 8.2.5 Changelog

8.2.4 - Jan 27, 2026

Issues Fixed:

• SERVER-112952 - tighten profile command auth check

• SERVER-113685 - Segfault for query on compound wildcard index

• SERVER-114126 - Reduce memory usage and add new guardrails to the boolean simplifier

• SERVER-114695 - Update default value for pending connections

• SERVER-114838 - Lock Manager is susceptible to an adversarial DDOS attack

• SERVER-115296 - Redact ResourceId raw values from error messages

• SERVER-116206 - Better exception handling for message compressors

• SERVER-116210 - Adjust the maximum buffer size for ingress requests

• SERVER-116211 - Cap memory allocated by network buffers

• All JIRA issues closed in 8.2.4

• 8.2.4 Changelog

8.2.3 - Dec 19, 2025

Issues fixed:

• SERVER-115508: Make minimally sized buffers for uncompressed Messages

• 8.2.3 Changelog

8.2.2 - Nov 19, 2025

Issues fixed:

• SERVER-103774: Disallow usage of compound wildcard indexes as shard key indexes

• SERVER-105047: Queries using $or and $ne on sharded collections might include orphan documents

• SERVER-110423: $group can use a distinct scan on a sparse index, which excludes results with null or missing values

• SERVER-110916: Resolve a use-after-move issue in targetShardsAndAddMergeCursors()

• SERVER-112467: Track and report the number of pending proxy connections

• SERVER-112520: Failure to write to a geo index when inserting a document into an existing timeseries bucket

• All JIRA issues closed in 8.2.2

• 8.2.2 Changelog

8.2.1 - Oct 3, 2025

Issues fixed:

• SERVER-94681: Move sampling to the oplog cap maintainer thread

• SERVER-106469: Starting mongocryptd fails on latest build, windows.

• SERVER-108565: Improve handling of oversized writes to time series collections.

• SERVER-109626: Adjust behavior for emitting additional fields in change streams

• SERVER-110250: Use existing session when dropping spill WT ident

• WT-15225: Fix newly created table and drop() EBUSY issue

• All JIRA issues closed in 8.2.1

• 8.2.1 Changelog

Geospatial Index Generation

Starting in MongoDB 8.2, if a document contains both GeoJSON and legacy numeric coordinate fields, geospatial index generation prioritizes GeoJSON coordinates.

If your documents contain legacy numeric coordinate values that precede GeoJSON coordinates and existing indexes rely on that behavior, you might need to rebuild your geospatial indexes.

This change might require you to rebuild geospatial indexes if your documents contain legacy numeric coordinate values that precede GeoJSON coordinates and existing indexes rely on that behavior. Review geospatial queries to confirm they return expected results.

terminateSecondaryReadsOnOrphanCleanup Parameter

MongoDB 8.2 adds the terminateSecondaryReadsOnOrphanCleanup parameter, which controls whether long-running read operations on secondary nodes automatically terminate before orphaned document deletion following a chunk migration.

By default, this parameter is set to true. If a read operation on a secondary node begins before the chunk migration commits, MongoDB automatically terminates the operation before deleting the orphaned documents. In MongoDB versions before 8.2, these operations continue executing after orphaned document deletion and may silently miss documents without returning an error.

To support this new behavior, the default value for orphanCleanupDelaySecs is now 3600 (previously 900).

Index Builds During Initial Sync

Starting in MongoDB 8.2, index builds during initial sync use 10% of available RAM by default. You can specify a different percentage of memory for MongoDB to use with the initialSyncIndexBuildMemoryPercentage parameter.

By default, the amount of available memory must be between 200 MB and 16 GB, inclusive. You use the following parameters to specify different minimum and maximum thresholds for memory usage:

• initialSyncIndexBuildMemoryMinMB

• initialSyncIndexBuildMemoryMaxMB

Query Statistics

Starting in MongoDB 8.2, the $queryStats output includes the following metrics on delinquent tickets:

• metrics.delinquentAcquisitions

• metrics.totalAcquisitionDelinquencyMillis

• metrics.maxAcquisitionDelinquencyMillis

MongoDB considers an execution ticket as delinquent when acquisition takes too long.

WiredTiger Cache Size Configuration as a Percentage

Starting in MongoDB 8.2, you can configure the WiredTiger internal cache size using a new percentage-based option. In addition to the existing --wiredTigerCacheSizeGB and storage.wiredTiger.engineConfig.cacheSizeGB gigabyte-based settings, you can now specify the cache size as a percentage of the available memory with the --wiredTigerCacheSizePct and storage.wiredTiger.engineConfig.cacheSizePct options.

New Inconsistency Type: Range Deletion Missing Shard Key Index

Starting in MongoDB 8.2, a new inconsistency type RangeDeletionMissingShardKeyIndex is implemented. The inconsistency type indicates that a sharded collection exists that doesn't have an index compatible with the collection shard key and there is at least one remaining range deletion task to complete.

New Aggregation Expression

Starting in MongoDB 8.2, you can use $currentDate with aggregate() to return the current time on the server.

New $currentOp Field

The $currentOp aggregation stage now includes the versionContext field, which provides information on the Feature Compatibility Version (FCV) of the operation.

serverStatus Output Change

serverStatus includes the following new fields in its output:

• metrics.operation.killedDueToRangeDeletion

• metrics.query.bucketAuto.spilledBytes

• metrics.query.bucketAuto.spilledDataStorageSize

• metrics.query.bucketAuto.spilledRecords

• metrics.query.bucketAuto.spills

• planCache.classic.replanned_plan_is_cached_plan

• planCache.sbe.replanned_plan_is_cached_plan

• spillWiredTiger

• connections.queuedForEstablishment

• connections.establishmentRateLimit

• connections.establishmentRateLimit.rejected

• connections.establishmentRateLimit.exempted

• connections.establishmentRateLimit.interruptedDueToClientDisconnect

• metrics.network.averageTimeToCompletedTLSHandshakeMicros

• metrics.network.averageTimeToCompletedHelloMicros

• metrics.network.averageTimeToCompletedAuthMicros

• queues.ingressSessionEstablishment

• queues.ingressSessionEstablishment.addedToQueue

• queues.ingressSessionEstablishment.removedFromQueue

• queues.ingressSessionEstablishment.interruptedInQueue

• queues.ingressSessionEstablishment.rejectedAdmissions

• queues.ingressSessionEstablishment.exemptedAdmissions

• queues.ingressSessionEstablishment.successfulAdmissions

• queues.ingressSessionEstablishment.attemptedAdmissions

• queues.ingressSessionEstablishment.averageTimeQueuedMicros

• queues.ingressSessionEstablishment.totalAvailableTokens

Explain Output for Disk Spilling Stages

Starting in MongoDB 8.2, the Explain Results for stages that spill to disk now include standardized metrics under consistent field names. This update provides a more unified way to track disk usage for memory-intensive operations.

The new fields include:

• spills

• spilledBytes

• spilledRecords

• spilledDataStorageSize

Repair Mode Field in Validation Output

Starting in MongoDB 8.2, the validate command and the db.collection.validate() helper method return a new repairMode field that indicates what types of data inconsistencies the validate command attempted to repair, if detected.

Support for zstd Fast Compression

Starting in MongoDB 8.2, The storage.wiredTiger.engineConfig.zstdCompressionLevel setting supports negative values for fast zstd compression. Negative values provide faster compression and decompression speeds at the cost of the compression ratio. The new supported range is -7 to 22.

Targeted Mirrored Reads

Starting in MongoDB 8.2, you can selectively mirror read operations to specific servers that need their caches warmed up by tagging the nodes for read mirroring. Unlike general mirrored reads, targeted read mirroring allows you to target hidden nodes and mirror from both primary and secondary nodes.

You can configure targeted mirrored reads using the targetedMirroring field in the mirrorReads parameter.

Rate Limit Ingress Connection Establishment

Starting in MongoDB 8.2, you can enable rate limiting for ingress connections to preserve CPU resources during overload. When enabled, you can see when the rate of incoming connections exceeds the connection establishment rate they specified and configure parameters to specify how your MongoDB deployment handles excess connections.

You can configure rate limiting using the following new fields:

• ingressConnectionEstablishmentRateLimiterEnabled

• ingressConnectionEstablishmentRatePerSec

• ingressConnectionEstablishmentBurstCapacitySecs

• ingressConnectionEstablishmentMaxQueueDepth

Database Profiler Parameters and Metrics

MongoDB 8.2.6 adds two parameters that let you control the impact that the Database Profiler has on query performance:

• internalQueryGlobalProfilingLockDeadlineMs

• internalProfilingMaxAbandonedWritesPerSecondPerDb

MongoDB 8.2.6 also adds two related serverStatus metrics:

• profiler.totalAbandonedWrites

• profiler.dbsPastThreshold

Queryable Encryption Prefix, Suffix, and Substring Query Public Preview

MongoDB 8.2 lets you enable prefix, suffix, and substring queries on encrypted string fields in Queryable Encryption enabled collections. For details, see enable substring queries and Supported Aggregation Expressions.

To use the Public Preview prefix, suffix, or substring queries with mongosh, you must separately download the Automatic Encryption Shared Library 8.2 or higher, then specify the library path to mongosh using the --cryptSharedLibPath option.

ingressRequestRateLimiterApplicationExemptions Parameter

Starting in MongoDB 8.2.2, the ingressRequestRateLimiterApplicationExemptions parameter lets you specify applications to exempt from ingress request rate limiting. For details, see ingressRequestRateLimiterApplicationExemptions.

Performance Improvements

MongoDB 8.1 introduces performance improvements from MongoDB 8.0, including but not limited to:

• Up to 195% improved throughput for time-series bulk insertions.

• Up to 40% improved throughput for match filter queries.

• Up to 20% improved throughput for queries on documents with arrays.

• Up to 10% improved throughput for in-cache read workloads.

• Up to 5% reduction in CPU utilization.

Add Comments to Query Settings

Starting in MongoDB 8.1 (and 8.0.4), you can use setQuerySettings to add comments to query settings. For example, add a comment that indicates why you added query settings.

Aggregation Accumulators

Starting in MongoDB 8.1, the following aggregation accumulators are available:

• $concatArrays

• $setUnion

Aggregation Stages

Starting in MongoDB 8.1, the following aggregation stages are available:

• $listClusterCatalog

Audit Log Improvements

Starting in MongoDB 8.1, if a client application connects to a mongod or a mongos instance through a load balancer, the origin client computer and load balancer IP addresses and ports are included in the audit log. You can use the log to match an audit event with the origin client computer.

For details, see OCSF Schema Audit Messages and mongo Schema Audit Messages.

connectionStatus Connection UUID

Starting in MongoDB 8.1, you can access the currently connected client's UUID through connectionStatus.authInfo.UUID.

Disk Space

Starting in MongoDB 8.1 (and 8.0.5), if disk space is running low, MongoDB will fail queries that are spilling to disk.

$geoNear Improvements

Starting in MongoDB 8.1, some previously failing $geoNear queries with hidden 2d or 2dsphere indexes will now succeed. In previous releases, some $geoNear queries used to unnecessarily fail with an IndexNotFound error when they included a hidden index due to confusion over which index to use.

Improved Accuracy for Operation Duration Metric

Starting in MongoDB 8.1, the durationMillis metric reported in slow query logs accounts for time spent processing authorization and parsing the command. As a result, durationMillis more closely reflects the full command duration.

Inconsistency Types

Starting in MongoDB 8.1, the following inconsistency types are available:

• CollectionAuxiliaryMetadataMismatch

$lookup Support for Encrypted Collections

Starting in MongoDB 8.1, you can reference multiple encrypted collections in a $lookup stage.

To learn more, see:

• $lookup with Encrypted Collections

• Supported Operations for Automatic Encryption

• Supported Operations for Queryable Encryption

$merge Missing Fields

Starting in MongoDB 8.1, if the $merge aggregation stage's supporting index is not sparse, the fields specified for the on option can be missing or contain a null value.

Query Statistics

Starting in MongoDB 8.1, query statistics are collected and reported for count and distinct commands. For details, see count Command Query Shape and distinct Command Query Shape.

Server Parameters

Starting in MongoDB 8.1, the following server parameters are available:

• enableAutoCompaction

• upsertMaxRetryAttemptsOnDuplicateKeyError

• AbortExpiredTransactionsSessionCheckoutTimeout (also available in 8.0.13)

• JWKSMinimumQuiescePeriodSecs

• catalogCacheCollectionMaxEntries

• catalogCacheDatabaseMaxEntries

• catalogCacheIndexMaxEntries

• wiredTigerSessionMax

serverStatus Metrics

Starting in MongoDB 8.1 (and 8.0.4, 7.0.14, and 6.0.20), the indexStats section of the serverStatus command output tracks indexes in prepareUnique state.

MongoDB 8.1 adds the following server status metrics:

• maxTimestampEligibleForTruncate in changeStreamPreImages.purgingJob

• metrics.ttl.invalidTTLIndexSkips

• metrics.repl.timestamps

• metrics.repl.timestamps.oldestTimestamp

• query.planning.fastPath.express

• query.planning.fastPath.idHack

• metrics.abortExpiredTransactions.successfulKills (also available in 8.0.13)

• metrics.abortExpiredTransactions.timedOutKills (also available in 8.0.13)

• wiredTiger.version

Slow Query Metrics for Disk Spilling

Starting in MongoDB 8.1, log messages for slow queries contain new metrics if the query execution writes temporary files to disk. These metrics are prefixed by the query execution stage that caused the query to exceed the memory limit. For example, sortSpills indicates the number of times that the sort stage of query execution wrote temporary files to disk.

For more information on writing temporary files to disk, see allowDiskUse().

Stable API

Starting in MongoDB 8.1, the renameCollection command and corresponding shell method, db.collection.renameCollection(), are included in Stable API V1.

Support for Search Stages in Explain Results

Starting in MongoDB 8.1, the explain results include execution stats for the $search, $searchMeta, and $vectorSearch stages.

To learn more, see:

• Execution Plan Statistics for Search Queries

• Explain Search Results

• Explain MongoDB Vector Search Results

Support for Search Index Commands on Views

Starting in MongoDB 8.1, you can run createSearchIndexes, updateSearchIndex, dropSearchIndex, and $listSearchIndexes to create, modify, drop, and list Atlas Search indexes and Atlas Vector Search indexes on compatible standard views that contain only the following stages:

• $addFields

• $set

• $match wrapping an $expr operation

You can run aggregation pipelines that contain $search and $searchMeta stages on views that use a search index. You can also run aggregation pipelines that contain a $vectorSearch stage on views that contain a vector search index.

Validate Command Index Specifications Output

Starting in MongoDB 8.1, the validate command includes index specifications in the validate.indexDetails output field.

Validation Action for Handling Invalid Documents

Starting in MongoDB 8.1, you can set your schema's validationAction option to errorAndLog, in which MongoDB rejects any insert or update that violates the validation criteria and logs the error to the mongod log file.

For more information, see Choose How to Handle Invalid Documents.

Only Single-Version Downgrades are Supported

MongoDB only supports single-version downgrades. You cannot downgrade to a release that is multiple versions behind your current release.

For example, you can downgrade a 8.2-series to a 8.0-series deployment. However, further downgrading that 8.0-series deployment to a 7.0-series deployment is not supported.

Downgrade Policy

• Binary downgrades aren't supported for MongoDB Community Edition.

• You cannot downgrade your deployment's FCV to or from a minor release version of MongoDB.

• If you upgrade or downgrade your deployment's FCV, you cannot downgrade your Enterprise deployment's binary version without assistance from support.