NoSQL Databases:HBase

HBase

• created in 2007 at Powerset
• distributed column-oriented data store built on top of HDFS
• available under Apache Software License

Features

• non-relational and sparse
• distributed
• persistent
• opensource
• horizontal scalable
• multidimensional rather than 2-D(relational)
• schema-free
• modeled after google's BigTable
• the decentralized storage system
• sorted map or Key/Value store
• easy replication support
• written in Java
• simple API, etc.

Architecture

Region

• A subset of a table’s rows
• Horizontal range partitioning

Region Server

• Manages data regions
• Serves data for reads and writes

Master

• Responsible for coordinating the slaves
• Assigns regions, detects failures
• Administrative functionality [2]

Tables, Rows, Columns, and Cells

• Its basic unit is a column.
• One or multiple columns from a row that is addressed uniquely by a row key.
• A number of rows, in turn, form a table, and there can be many of them.
• Each of the columns may have distinct value contained in a separate cell.

All rows are always sorted lexicographically by their row key.[1]

Data Model

• Map indexed by a row key, column key and a timestamp
• supports lookups, inserts, deletes (single row transactions only)

Rows are composed of columns, and those, in turn, are grouped into column families. All columns in a column family are stored together in the same low-level storage file, called an HFile. Millions of columns in a particular column family. There is also no type nor length boundary on the column values.

Rows and columns in HBase

References:

1. "HBase:The Definitive Guide",Lars George,page-17
2. "Introduction to Big Data Analytics" at corsera.org,Natasha Balac
3. "Apache HBase Reference Guide" at hbase.apache.org