[Elasticsearch] This Week in Elasticsearch and Apache Lucene - 2016-04-25
Elastic/Elasticsearch 2016. 4. 26. 15:48이번 weekly 에서 눈에 확 들어 오는건 개인적으로 아래 두 가지 입니다.
- Thread local leaks when running in web containers have finally been fixed.
- CamelCase support has been removed.
원본 글)
https://www.elastic.co/blog/this-week-in-elasticsearch-and-apache-lucene-2016-04-25
Elasticsearch Core
Changes in 2.x:
- The index name was missing from the search slowlog.
- CamelCase is deprecated (and has deprecation logging).
- MoreLikeThis now handles aliases correctly.
Changes in master:
- The .percolator type has been replaced with the percolator field datatype.
- Added a fingerprint token filter and fingerprint analyzer for duplicate detection.
- TransportReplicationAction has been signficantly refactored in order to make it unit testable.
- RPM and Deb packages now set permissions explicitly, instead of relying on umasks.
- Indexed scripts and templates are now stored in the cluster state, and are called "stored" scripts/templates.
- Parameter names in ingest processors are now more consistent.
- IP fields support range queries again.
- readNamedWriteable and writeNamedWriteable are now public, and writable.readFrom is gone.
- UUID generators moved out of Strings, to avoid spooky action at a distance.
- The `action.realtime_get` setting has been removed.
- Support for unquoted JSON keys can be allowed via a system property, for bwc purposes.
- Cross-type mapping updates were not working for boolean fields.
- Empty task IDs are now serialised in 1 byte, so that every task can have a parent ID.
- Reindex child tasks weren't being marked as such.
- Validation failures have been removed from the cluster health response.
- Object fields now inherit their dynamic setting from their parent object or type.
- Thread local leaks when running in web containers have finally been fixed.
- Added a safeguard to protect against too-large rescore windows.
- The elasticsearch-plugin script now prints the download URL of the plugin when in verbose mode, and has friendlier error messages.
- The startup script now fails with an error code if the elasticsearch binary is not found or is not executable.
- CamelCase support has been removed.
- The ICU analyzer now accepts custom rule files.
Ongoing changes:
- Dots in fields names are now supported, but so far only if the parent fields already exist. Tests are being added to make sure supporting dots fully doesn't break anything.
- Persistence of results of long running tasks.
- A `minhash` token filter for estimating the Jacard similarity coefficient between two docs.
- Pipeline aggs are only needed on the coordinating node.
- Adding searchable/aggregatable tags to fields in the field stats API.
- Inner hits will no longer support the top-level syntax as the inline syntax has been improved.
- It should be possible to pass include/exclude values to the terms aggs using the same format that was used to render bucket keys.
- Deleted index tombstones close to being merged.