Hortonworks Installation

[#1]

Is the Hortonworks Data Platform installation/setup as outlined under

http://docs.hortonworks.com/HDPDocuments/HDP2/HDP-2.3.4/bk_installing_manually_book/content/ch_getti...

the standard way to setup Hortonworks Hadoop platform?

I can see that this is listed as the 'manual setup'.

So is the automated way to install Ambari and then setup the cluster through Ambari?

Appreciate the clarification.

[#2]

Yes, your understanding is correct. Automated and recommended way is to first install Ambari and then setup cluster through Ambari.

[#3]

In the manual installation as cited in my first post, where is the HDP part? It all seems to be the installation/setup of only Apache Hadoop and its suite of products (HDFS, MR, YARN, Zookeeper, HBASE, HIVE etc) ie I don't see anything really different from a standard Apache Hadoop install/setup.

In that manual installation sequence I don't see any install of Ambari either.

What is Hortonworks about the manual installation?

Appreciate the clarification.

[#4 · Accepted solution]

HDP is Apache Hadoop and its suite of products (HDFS, MR, YARN, Zookeeper, HBase, Hive etc.). In the manual install you don't need Ambari. That's why it's all manual. Ambari manages HDP that it installs itself. If you do manual, then you need to other tools to monitor and manage it.

When you do manual install and say "yum install hadoop hadoop-hdfs hadoop-libhdfs hadoop-yarn hadoop-mapreduce hadoop-client openssl", where do you think these packages are being installed from? It is the repo that you setup in the following step when you configure remote HDP repositories. So all this is actually part of HDP.

http://docs.hortonworks.com/HDPDocuments/HDP2/HDP-2.3.4/bk_installing_manually_book/content/config-r...

If you have a followup question, please add a comment instead of posting as a new answer.

[#5]

HDP stands for Hortonworks Distributed Platform. It refers to all components supported by Hortonworks. HDP Hadoop is actually certified by Hortonworks. If you want to use the Hortonworks certified Hadoop, I would recommend you to use Ambari to installer for installation.

Find the main two ways to install a HDP cluster.

1) HDP install via Ambari ( Recommended)

Ambari is the HDP stack installer. In this method, a user typically install ambari on all the hosts and use ambari to install HDP stack such as HDFS, MR, Zookeeper, HBase etc.

You can find the detail steps for ambari install at http://docs.hortonworks.com/HDPDocuments/Ambari-2.4.2.0/bk_ambari-installation/content/ch_Getting_Re... ( This is latest ambari version.)

2) Manually Install

This option means that user will install the all components manually by running "yum install" on each node.

Here, user will be responsible for installing all the necessary packages and set the configuration properly on all the hosts.

Hortonworks also provide a manual guide to install HDP components.

http://docs.hortonworks.com/HDPDocuments/HDP2/HDP-2.3.4/bk_installing_manually_book/content/ch_insta...

[#6]

Once the "Ambari" setup done, you can achieve all other thing in automated way.

[#7]

can any one please send me the ambari installation document on ubuntu