Hbase a fully open source and proven highly scalable distributed database.
Hbase presents a parallel column-oriented database built as the persistant storage engine for Hadoop Distributed Filesystem. Hbase provides easy integration to Hadoop and MapReduce. Scaling out is done by adding nodes to the cluster. For those reasons HBase functions as de-facto high performance storage engine for large dataset residing on HDFS. HBase is a highly scalable database that can deliver very high performance read-write operations on data-intensive applications.HBase offers quasi-partitioning implemented as sparsely populated tables with the ability to add columns during operation. A typical Hbase set-up will comprise a look-up table storing metadata describing the region (location) of given partitioned tables for access speed-up.Hbase offers granularity for analyzing data one record at a time extracted from the very large dataset in HDFS. It thus provide a very low latency solution to providing RDBMS-like capability for HDFS.
SOA and data federation.Federating data enabled by RESTful interface from distributed web application (SOA)
Built for data intensive web application. Using a REST web services integration and interfacing with non-java application, providing the transactional data querying and extraction speed , and low latency that limits HDFS.
