• • Linear and modular scalability.
  • Strictly consistent reads and writes.
  • Automatic and configurable sharding of tables.
  • Automatic failover support between RegionServers.

• • Convenient base classes for backing Hadoop MapReduce jobs with Apache HBase tables.
  • Easy to use Java API for client access.
  • Block cache and Bloom Filters for real-time queries.
  • Query predicate push down via server side Filters.
  • Thrift gateway and an REST-ful Web service that supports XML, Protobuf, and binary data encoding options.

How to use Apache HBase REST Interface

• Extensible JRuby-based (JIRB) shell.
• Support for exporting metrics via the Hadoop metrics subsystem to files or Ganglia; or via JMX.