The Elastic team released version 7.0 rc1 of its open source distributed, RESTful search and analytics engine, Elasticsearch, last week. Elasticsearch 7.0 rc1 explores new features, breaking changes, deprecations, and bug fixes among others.
New features
- Allocation: There’s a new Node repurpose tool in Elasticsearch 7.0 rc1.
- Security: Internal security index has been switched to “.security-7”.
Breaking Changes
- Distributed: Cluster state size has been removed.
- Features: Migration upgrade and Assitance APIs are removed.
Enhancements
- Retention lease sync intervals have been reduced.
- Retention leases have been integrated to recovery from remote.
- Dedicated retention lease exceptions have been added.
- Support has been added for selecting percolator query candidate matches in elasticsearch 7.0 rc1 that contains geo_point based queries.
Bug Fixes and deprecations
- Size has been deprecated in cluster state response
- Fallback has been deprecated to java on PATH
- The issue of sibling pipeline aggregators reduction during non-final reduce has been solved.
- nextDoc has been extended to delegate to the wrapped doc-value iterator for date_nanos.
- Non-super users are now allowed to create API keys.
- A consistent view of realms is now used for authentication.
- Reading auto-follow patterns from x-content has been enabled in Elasticsearch 7.0 rc1.
- auto-followers have been stopped on shutdown.
- Node tool cleanup has been fixed.
- Serializing state will be now avoided in case it is already serialized.
- waitForActiveShards will be ignored when syncing leases.
- multi_value_field_leniency has been added inside FieldHitExtractor.
For more information, check out the official Elasticsearch 7.0 rc1 release notes.
Read Next
Elasticsearch 6.5 is here with cross-cluster replication and JDK 11 support
Search company Elastic goes public and doubles its value on day 1