Apache Lucene and Solr 5 released! | Solr Enterprise Search
Some of the changes in Lucene library: File access from Lucene is always done using Java NIO.2 classes, which results in safer and more reliable library Each Lucene segment is given its own identifier making replication and its verification easier IndexWriter checks segments checksum before performing segments merge operation Heap memory usage has been lowered especially during segments merging thanks to new codec – Lucene50Codec FieldCache has been removed and moved to special class – sorting should be done on doc values fields from now New field type has been introduced – DateRangeField. It allows date range searches on multivalued fields ConcurrentMergeScheduler is now able to detect if its running on SSD drive and adjust itself is that's true Added the possibility to choose between better compression or better performance for stored fields Some of the changes in Solr search engine:Read full article from Apache Lucene and Solr 5 released! | Solr Enterprise Search
No comments:
Post a Comment