2.1.0 release highlights

edit

New and notable

edit

New and notable changes in version 2.1.0 of Elastic Cloud on Kubernetes. Check Elastic Cloud on Kubernetes version 2.1.0 for the full list of changes.

Improved Elasticsearch status sub-resource

edit

Additional information was added to the Elasticsearch status sub-resource, which provides rich details concerning the in-progress operations during upgrades, upscale, and downscale operations. New conditions fields include ReconciliationComplete, RunningDesiredVersion, and ElasticsearchIsReachable which gives information explaining why each condition is either True, or False. Also included is a new parent field inProgressOperations, which provides topology information for upgrades, upscale, and downscale operations.

Improved Elasticsearch and Kibana generation status

edit

An additional field observedGeneration is now maintained within Elasticsearch and Kibana’s status sub-resource. This new field represents the current generation of the specification that the ECK operator is working to reconcile, and is intended to allow tools to deterministically monitor the rollout of custom resources.

Allowing upgrade predicates to be selectively disabled

edit

Starting with ECK 2.1, the Elasticsearch clusters can have certain upgrade predicates (rules) disabled on a case-by-case basis using annotations on the Elasticsearch custom resource, which allow full control over what rules are considered during the Elasticsearch upgrade process. Selectively disabling the predicates is extremely risky, and carries a high chance of either data loss, or causing a cluster to become completely unavailable. This feature is therefore intended exclusively as a troubleshooting mechanism of last resort. Check the documentation for more details.

Known issues

edit
  • Under certain circumstances the operator will keep terminating and restarting Elasticsearch Pods seemingly at random. The underlying issue is fixed in ECK 2.4.0 and an upgrade is highly recommended.