(Apache) Hadoop

About

Apache Hadoop is basically 3 things:

Yarn Hortonworks Image comes from Hortonworks

Hadoop:

  • scales, runs on commodity hardware where you setup a Hadoop cluster (a group of machines connected together over a network)
  • is fault tolerant.
  • is good for “embarassingly parallel” problems, such as building an index for the web.

Steps:

  • setting up a Hadoop cluster and then store huge amounts of data spread across all the machines into the HDFS
  • processing the scattered data by writing MapReduce programs (or jobs) that goes locally on each machine in order to not relocate the data.

Single threaded writes.

Use cases

  • Data Warehouse/Analytics Offload
  • Data Lake/Enterprise Data Hub

Building

The BUILDING.txt in the code repository has all information to build it (also for windows)

Proposition

Oracle: Big Data Appliance

In Oct 2011, Oracle announced the Big Data Appliance, which integrates:

  • Cloudera's Distribution Including Hadoop (CDH),
  • Oracle Enterprise Linux,
  • the R programming language,
  • and a NoSQL database

with the Exadata hardware.

Azure

Azure - HDInsight (Microsoft's Hadoop)

Documentation / Reference

Task Runner