danburkert / kiji-schema

A simple Java API and command line interface for importing, managing and retrieving data from HBase.

KijiSchema

KijiSchema is a layout and schema management layer on top of Apache HBase. KijiSchema supports complex, compound data types using Avro serialization, as well as cell-level evolving schemas that dynamically encode version information.

KijiSchema includes both commandline tools and a simple Java API to manipulate tables, layouts, and data. It also provide native MapReduce Input/Output formats for Kiji tables.

For more information about KijiSchema, see the Kiji project homepage.

Compilation

KijiSchema requires Apache Maven 3 to build. It may be built with the command

mvn clean package

or by executing bin/kiji from the root of a checkout. This will create a release in the target directory.

Installation

KijiSchema requires an HBase and Hadoop cluster to run. Confirm that $HADOOP_HOME and $HBASE_HOME are set appropriately. Then issue the following command from your KijiSchema root directory:

bin/kiji install

This will install a Kiji instance on the hbase cluster named 'default'.

Usage

For commandline tool usage, issue the command:

bin/kiji

Further documentation is available at the Kiji project Documentation Portal