Please use the form below to provide your feedback. Because your feedback is valuable to us, the information you submit in this form is recorded in our issue tracking system (JIRA), which is publicly available. You can track the status of your feedback using the ticket number displayed in the dialog once you submit the form.

What’s New?

      +

      Couchbase Kubernetes Operator 2.9 was released in December 2025. New features and improvements are described below.

      For information about fixed and known issues, see the Release Notes.

      New Features and Enhancements in 2.9

      Couchbase Kubernetes Operator 2.9 adds support for Couchbase Server 8.0 features and introduces a circuit breaker called Manual Intervention Required (MIR) mode. This release also improves the upgrade process by introducing a new upgrade object that enables more resilient automated upgrades.

      Support for Couchbase Server 8.0 Features

      Couchbase Server 8.0 introduces several key features, listed in What’s New in Version 8.0, which are now configurable through Couchbase Kubernetes Operator 2.9. These are the highlights:

      • Encryption at Rest: Added support for a new CouchbaseEncryptionKey custom resource to manage encryption keys and to specify which keys are used for each unit, such as buckets, configuration data, and logs.

      • Magma as the Default Storage Engine: In Couchbase Server 8.0, Magma with 128 vBuckets is the default storage engine for new buckets. Therefore, for CouchbaseBuckets, now the default storage engine is magma and the vBucketCount is 128. Couchbase recommends setting the vBucket count to 1024 for high-throughput workloads.

      • XDCR Conflict Logging for Active-Active Setups: Added multiple settings to the CouchbaseBucket resource to set up XDCR Conflict Logging.

      • Other features such as enabling and disabling user accounts, and new bucket settings such as for warmup.

      Manual Intervention Required (MIR) Mode

      The MirWatchdog is an out-of-band check that provides additional alerting. It’s used when the Operator cannot reconcile a cluster due to reasons outside its control and requires manual user intervention. Scenarios include TLS expiration, Couchbase authentication errors, and rebalance failures.

      This feature is disabled by default but can be enabled and configured by using the mirWatchdog field in the CouchbaseCluster CRD.

      For more information, see Monitor for Manual Intervention Scenarios.

      Upgrade Process Improvements

      Autonomous upgrades now provide improved user input and control through a new upgrade object in the CouchbaseCluster specification, which contains all configurations for upgrading a cluster. The previous ServerClass image-based approach is now hard deprecated.

      This consolidation clarifies how to control and manage future upgrades. All existing upgrade fields upgradeProcess, upgradeStrategy, and rollingUpgrade are deprecated, and moved under the upgrade object.

      The new upgrade object has the following new fields:

      • UpgradeOrderType: The unit of upgrade that can be a Node, ServerGroup, ServerClass, or Service.

      • UpgradeOrder: The upgrade order of specified upgrade units.

      • stabilizationPeriod: The wait time in seconds before upgrading the next unit.

      • previousVersionPodCount: The number of pods to keep running on the previous version at the end of upgrade.

      For more information, see Upgrade Couchbase Server and Upgrade a Couchbase Deployment.

      Other Improvements

      The following are additional important improvements:

      • Support for changing the bucket eviction policy online without requiring a bucket restart.

      • The ability to disable DNS resolution verification when creating pods before activating them in the cluster.

      • Support for specifying overheadMemory for autoResourceAllocation to define a static overhead amount.

      • Operator-backup now supports Periodic Merge in addition to Full Only and Full Incremental. This strategy is better suited for some larger clusters. You can specify a merge schedule on the CouchbaseBackup resource.

      • Support for specifying the cao.couchbase.com/additionalArgs annotation on CouchbaseBackup and CouchbaseRestore resources to pass additional cbbackupmgr arguments to the container.

      More Information