The search is over
Lucene and Solr reach version 4.0
years in the making, today sees the release of version 4.0 of
open-source search engine Solr and underlying Java library
Originally created by Doug Cutting (who has since gone on to create Hadoop) in 1999, Lucene is now used to power Apple’s online stores among many other applications and sites. The Lucene project’s key components are Lucene Core, a Java library of indexing and search functions, and Solr, the search server built using Lucene Core.
Solr’s most exciting new features come under the banner of SolrCloud, including distributed indexing which allows multiple servers to carry out indexing at once. This increase in speed allows for continuous indexing and “near real-time” searching, which is for many a baseline expectation of search engines in 2012. Distributed indexing also allows for NoSQL features such as realtime-get, optimistic locking, and durable updates.
Another aspect of SolrCloud is the integration of another Apache project, Zookeeper, for distributed coordination and to provide immunity to split-brain issues.
Lucene Core, meanwhile, has seen a large number of technical additions to its library, which can be found in its release notes.
This latest version promises to be considerably most more reliable than most X.0 releases, since the project now releases formal Alpha and Beta versions, months in advance of the GA, for thorough testing.
You can download the 4.0 releases of Lucene and Solr now.