README

AccumuloBook QuickInstall

******************************************************************************
0. Introduction


This project creates a tar.gz that includes Accumulo and all dependencies. The
archive file unpacks into one directory and sets up Accumulo to run on one
node.  The only requirement to run the tar.gz is Java 1.7.  To build the
tar.gz, Maven 3.0.4 or greater is required.  You will need at least 4G of
RAM to run reasonably.  More is better.


******************************************************************************
1. Installation

There are three steps to install the AccumuloBook QuickInstall.

    - Download the file from
      https://accumulobook.com/quickinstall

    - Extract that file somewhere, it will expand into a quickinstall-home
      directory

    - Change directory to quickinstall-home and run

       ./bin/install

******************************************************************************
2. Usage

After running the install, Hadoop, Zookeeper and Accumulo should be running.

To setup your path run

    source bin/quickinstall-env

To stop Accumulo, Zookeeper and Hadoop run

    qi-stop

To start it up again, run

    qi-start

There will also be a couple of shortcut commands to open the local docs.
These pages link to local copies of the documentation and API docs.  Run

    accumulo-doc

or

    hadoop-doc

Here are some useful links.
  The Accumulo monitor page should be available at http://localhost:50095
  HDFS should be available at http://localhost:50070
  The Yarn resource manager should be available at http://localhost:8088

The default user/password setup for accumulo is root/secret.  To change it run

    accumulo shell -u root -p secret

to get into the shell.  Then run the following to change it.

    passwd -u root


******************************************************************************
3. Note about native libraries

Hadoop comes packaged with native libraries for 32 bit Linux.  Native libraries
for 64 bit linux have been compiled following the instructions at
http://hadoop.apache.org/docs/r2.4.1/hadoop-project-dist/hadoop-common/NativeLibraries.html
Native libraries for Mac OSX have been built following instructions at
http://gauravkohli.com/2014/09/28/building-native-hadoop-v-2-4-1-libraries-for-os-x

The native library for Hadoop that are appropriate for your system are symlinked to
HADOOP_HOME/lib and then configured in Hadoop and Accumulo.

Accumulo native libraries built using ACCUMULO_HOME/bin/build_native_library.sh.
If the script is successful, sample configs are copied from
ACCUMULO_HOME/config/examples/2GB/native-standalone.  If the native libraries
fail to build, sample configs are copied from
ACCUMULO_HOME/config/examples/2GB/standalone

If for any reason native libraries are not found or built, both Accumulo and
Hadoop will run just fine.  They both have Java libraries that handle the same
functionality as the native libraries, they are just slower.


******************************************************************************
4. Building the tar.gz


To build the distribution, you will need Apache Maven installed.  Version
3.0.4 was used in development of this package.  Run the following command to
create an accumulo-quickinstall-1.6.1.tar.gz file in the target directory that
you can copy wherever you want to install the quickstart.

    mvn clean package

We use the maven-download-plugin from googlecode to download the binary
distributions of Hadoop, Zookeeper and Accumulo.  The first time you
build this, it will take some time to download all the components.

After the first download, these files are cached in your ~/.m2 directory.
Subsequent builds will not require they be downloaded again.