Elasticsearch 2.0.0 lands, bringing numerous new features
The blockbuster 2.0.0 release of beloved full-text search engine Elasticsearch has landed. Based on Lucene 5.2.1, Elasticsearch 2.0.0 GA has countless changes for increased productivity, as well as more matured security features, not to mention the new ‘pipeline aggregation’.
Following a beta-release sneak preview in September, Elasticsearch 2.0.0 has arrived, and with it a big security improvement in the form of the Java Security Manager. This tool makes it difficult for hackers to gain access to Elasticsearch applications.
Alongside numerous minor changes in performance and reliability, the release also contains a new “pipeline aggregation”, which lets users create aggregations based on those that already exist. Currently for testing only and not for production, the sneak peak release comprises more than 2,500 pull requests from over 400 committers.
The search database of choice for brands like LinkedIn, eBay and WordPress, Elasticsearch has made a concerted effort to deliver better data compression while also reducing memory consumption during merges. The command structure of the query execution has been optimised, and an auto-caching of filters is now also ready for use.
Instead of auto-regulatory settings, the new version focuses on feedback loops. The developer team have cleaned up the type mappings to make them “safe, unambiguous and reliable”. Additional features and details on the release can be found on the Elasticsearch blog.
“Filters are no more”
A major highlight in the new release is the merging of filters and queries. All filters clauses in Elasticsearch are now query clauses, and according to Elasticsearch, this means that the “execution can be automatically optimized to run in the most efficient order possible.”
The Elasticsearch team also announced it will take a new approach in the development of officially supported core plugins. These plug-ins are now in the Elasticsearch repository on GitHub. They are being tested in parallel with the core of new versions and being released for use. Make sure to check out the specific documentation for adding core plug-ins, which can be installed with Shield or Watcher.
SEE ALSO: A brief history of Elasticsearch
The second major release of Elasticsearch is based on Lucene 5.2.1. With its most recent release announced in August, Apache Lucene 5.3.0 has brought a number of improvements to its spatial3d module, which allows to make requests in a three-dimensional space. A new Lucene codec also makes it easier to save resources when indexing geo-objects like bounding boxes or polygons.
For an overview of all new features and changes, check out the Elastic-blog. The latest version is already available to download from the company website, and the migration plugin is on GitHub, for anyone experiencing trouble when upgrading.