Apache Cassandra reaches foretold version 1.0
The Apache Software Foundation (ASF) has announced the release of version 1.0 of the open source, highly scalable, column-oriented, distributed "NoSQL" database, Cassandra. The release comes just under five months after the release of the previous version, 0.8.0, and since then the developers have added support for data compression to reduce the volume of data on disk on Cassandra nodes and have improved the memory and disk space management with off-heap storage of the row cache and self tuning memory tables.
Other enhancements include the use of leveled compaction based on the strategies in Google's LevelDB, support for running Cassandra as a Windows service, multi-threaded compaction and easier node management and failover. According to a posting by Cassandra specialists Datastax, over the last year (since version 0.6), read performance has increased by more than 400 per cent and write performance by 40 per cent.
Cassandra is a distributed database management system that was originally developed by Facebook; it was open sourced in July 2008 to become the basis of a project in the Apache Foundation's Incubator. In late February 2010, it was accepted as an Apache Top-Level Project (TLP). It is used by companies such as Digg, Twitter, Rackspace, Cisco, Netflix and Reddit. According to the ASF, the largest known production cluster of Cassandra carried over 300 terabytes of data spread over 400 machines.
Details about Cassandra 1.0 can be found in the mailing list announcement, the release notes and in the change log. Cassandra 1.0 is available to download from the project's web site. Apache Cassandra is licensed under the Apache Licence 2.0. The 0.8 branch of Cassandra, currently at version 0.8.7, will continue to be maintained, for a while, after the release of 1.0.