v2.10.0

MongoDB Atlas Operator 2.10.0

Improvements and fixes

• Adds support for in-place upgrades from Flex to Dedicated clusters directly through the Atlas Kubernetes Operator. For more information, see Flex to Dedicated migration guide. #2526 by @helderjs
• Updates DEBUG logging level to include detailed differences of resources that the Atlas Kubernetes Operator changes. For more information, see Log Resource Diffs. #2525 by @igor-karpukhin

Images can be found in: https://quay.io/mongodb/mongodb-atlas-kubernetes-operator

Supported Kubernetes versions updated in this release: 1.31 - 1.33
Supported OpenShift versions updated: 4.18
See our full compatibility matrix here.

v2.9.1

MongoDB Atlas Operator 2.9.1

Improvements and fixes

• Fixed a bug with missing RBAC for the AKO 2.9.0 on Openshift causing AKO not able to create AtlasThirdPartyIntegration resources in Atlas. #2421 by @josvazg
• Fixed a bug that could stop AKO from syncing a resource to Atlas if an AtlasThirdPartyIntegration was created or modified. #2422 @josvazg
• Fixed a bug that could cause AtlasThirdPartyIntegration to repeatedly update in a loop when importing existing integrations in Atlas. #2425 by @josvazg

Images can be found in: https://quay.io/mongodb/mongodb-atlas-kubernetes-operator

Supported Kubernetes versions updated in this release: 1.30 - 1.32
Supported OpenShift versions updated: 4.18
See our full compatibility matrix here.

v2.9.0

MongoDB Atlas Operator 2.9.0

Improvements and fixes

• Adds the Atlas Third Party Integrations independent custom resources. Configuring these resources via the AtlasProject CRD is now deprecated. Support will be removed in a later release of the Operator. Please consult the migration guide. #2399 by @josvazg
• Fixes a bug that prevents usage of GCP region configurations for Atlas for Government deployments. The fix removes government region and provider restrictions from AtlasDeployment regionConfig validation. You can now deploy and manage Atlas resources on GCP without encountering validation errors or configuration limitations. #2367 by @helderjs

Images can be found in: https://quay.io/mongodb/mongodb-atlas-kubernetes-operator

Supported Kubernetes versions updated in this release: 1.30 - 1.32
Supported OpenShift versions updated: 4.18

See our full compatibility matrix here.

v2.8.2

MongoDB Atlas Operator 2.8.2

Improvements and fixes

• Fix a bug where the operator infinitely reconciles AtlasDeployment custom resource because of incorrect region configs comparison. This happens if the compute autoscaling is set to false, but minInstanceSize and maxInstanceSize fields are set. #2347 by @s-urbaniak
• Fixed an issue where the expired AtlasDatabaseUser resources couldn’t be deleted #2343 by @s-urbaniak
• Fixed a bug where users couldn’t set tags with spaces for AtlasDeployment resources #2330 by @igor-karpukhin
• Fixed an issue where backup compliance policies were not able to be created if no on-demand policy has been specified #2331 by @tvdw-mongodb
• Fixed an issue when AKO defaulted mongodb version to 7.0 for AtlasDeployment custom resource, if MongoDBMajorVersion field is not set by a user. With this fix, AKO doesn’t default to any version. If this field is not set, Atlas will set it to the latest major version. #2349 by @igor-karpukhin
• Added missed field “Description” for the AtlasDatabaseUser resource #2357 by @tvdw-mongodb

Images can be found at: https://quay.io/mongodb/mongodb-atlas-kubernetes-operator

Supported Kubernetes versions: e.g. 1.30-1.32
Supported OpenShift versions: e.g 4.17

v2.8.1

MongoDB Atlas Operator 2.8.1

Improvements and fixes

• Fixed an issue causing an infinite reconciliation loop when the scaleDownEnabled flag was disabled but the minInstanceSize field was not cleared. #2208 by @josvazg
• Fixed an issue where the project could remain in an unready state due to incorrect status conditions for unmanaged private endpoints. #2236 by @helderjs
• Fixed an issue with dry-run not capturing all errors related to teams. #2268 by @helderjs
• Updated golang.org/x/oauth2 to address security vulnerability CVE-2025-22868. #2196 by @dependabot
• Updated golang.org/x/net to address security vulnerability CVE-2025-22870. #2191 by @dependabot

Images can be found in: https://quay.io/mongodb/mongodb-atlas-kubernetes-operator

Supported Kubernetes versions updated in this release: 1.30-1.32
Supported OpenShift versions updated: 4.17

v2.8.0

MongoDB Atlas Operator 2.8.0

New features

• PUBLIC PREVIEW: Added dry run functionality , allowing users to preview changes to Atlas resources after making a change in AKO. The primary use case for this is to test an upgrade to the AKO. #2051 by @igor-karpukhin and @s-urbaniak
• Atlas Network Peerings and Network Containers can now be configured as independent custom resources. Configuring these resources via the AtlasProject CRD is now deprecated, and support will be removed in a later release of the Operator. Please consult the migration guide. #2100 & #2103 by @josvazg

Note

Starting with the 2.8.0 Atlas Kubernetes Operator release, the Kubernetes Atlas CLI commands are now available as a separate plugin, available at https://github.com/mongodb/atlas-cli-plugin-kubernetes. This plugin will be installed automatically, whenever atlas kubernetes CLI commands are executed. Updating the plugin to the newest version can be done using the atlas plugin update command.

Improvements and Fixes

• Fixed an issue where connection secrets for Flex clusters would be incorrectly cleaned up when multiple clusters had connections secrets in the same namespace. #2141 by @roothorp
• Fixed an issue where reconciling Backup Compliance Policies could lead to a nil panic. #2162 by @segui-mdb
• Correctly allow Serverless alerts to specify decimal thresholds for alerting conditions. #2160 by @tvdw-mongodb

Images can be found at: https://quay.io/mongodb/mongodb-atlas-kubernetes-operator

Supported Kubernetes versions: 1.30 - 1.32
Supported OpenShift versions: 4.17

v2.7.1

MongoDB Atlas Operator 2.7.1

Improvements and fixes

• Fixed a bug blocking creation of Flex clusters when trying to create a new Serverless instance. #2107 by @s-urbaniak
• Fixed a bug where connection secrets are not generated for Flex clusters when a database user is added. #2119 #2121 by @roothorp
• Fixed a bug in AtlasDatabaseUser resources that blocked unsetting the scope once added in the spec. #2120 by @helderjs
• Fixed a bug where applying an AtlasProject may delete network peering that is not managed by (not configured in) the AtlasProject custom resource. #2086 #2101 by @Josvaz

Images can be found at: https://quay.io/mongodb/mongodb-atlas-kubernetes-operator
Supported Kubernetes versions: 1.30 - 1.32
Supported OpenShift versions: 4.17

v2.7.0

MongoDB Atlas Operator 2.7.0

New features

• Support Flex clusters as the successor of the deprecated M2, M5 and Serverless clusters #2030 by @roothorp
• Atlas Project IP Access Lists are now configurable as independent custom resources. This allows for accidental deletion protection. IP Access list configuration within a Project custom resource is now deprecated, and support will be removed in a later release of the Operator. #2027 #2063 by @helderjs

Improvements and fixes

• Fix a bug where the operator could crash if the ReplicationSpec field in an `AtlasDeployment` custom resource is set but left empty, the ReplicationSpec field has empty entries, RegionConfigs are set but empty, or RegionConfig entries are empty. #2073 by @s-urbaniak
• Fixed an issue where reconciliation of private endpoints (used as independent custom resources) happens only every 3 hours instead of every 15 minutes #2068 by @helderjs
• Fixed an issue where AKO ignores the migration of existing custom roles maintained as AtlasProject subresources to independent custom resources. #2094 by @igor-karpukhin

Deprecations

• M2, M5, and serverless instances are deprecated. Support for these instances will be removed in a future release.

Images can be found in: https://quay.io/mongodb/mongodb-atlas-kubernetes-operator
Supported Kubernetes versions updated in this release: 1.28-1.30
Supported OpenShift versions updated: 4.17

v2.6.1

MongoDB Atlas Operator 2.6.1

Bug fixes

• Fixed a bug where Custom Roles configured as part of AtlasProject resources were deleted #2023 by @igor-karpukhin
• Fixed a bug that limits maintaining only up to 100 Data Federation, IP Access List, Network Peering, Teams, Federated Authentication, Alert Configurations, and Stream Connection resources #2016 by @s-urbaniak

Images can be found at: https://quay.io/mongodb/mongodb-atlas-kubernetes-operator

Supported Kubernetes versions: 1.28-1.30
Supported OpenShift versions: 4.16

v2.6.0

MongoDB Atlas Operator 2.6.0

New features

• Added support for the Workload/Workforce Identity Federation feature #1969 by @roothorp
• Private Endpoints can now be optionally configured as an independent custom resource - AtlasPrivateEndpoint #1879 by @helderjs
• Custom Roles can now be optionally configured as an independent CR - AtlasCustomRole #1913 by @igor-karpukhin

Improvements and fixes

• Fixed a bug when database user secrets were deleted for Serverless deployments as they were accidentally treated as orphaned secrets #1955 by @josvazg
• Resolved a case when AKO was not able to create AtlasSearch indexes #1956 by @igor-karpukhin
• Fixed incorrect reconciliation for DataFederation private endpoints. Prevents accidental deletion of data federation private endpoints that are not tracked by AKO #1929 by @s-urbaniak
• Fixed incorrect internal comparison for Atlas Project integrations that lead to not reconciling changes to Integrations #1902 by @josvazg
• Fixed a bug when the DataFederation controller was not watching AtlasProjects resources. Changes to AtlasProject resources were ignored which could lead to inconsistent state in AtlasDataFederation resources. This is fixed in #1868 by @s-urbaniak

Images can be found in: https://quay.io/mongodb/mongodb-atlas-kubernetes-operator

Supported Kubernetes versions updated in this release: 1.28-1.30
Supported OpenShift versions updated: 4.16