Understanding Big Data and Hadoop 

Introduction to big data, limitations of existing solutions

Hadoop architecture, Hadoop components and ecosystem

Data loading & reading from HDFS

Replication rules, rack awareness theory

 Hadoop cluster administrator

 Roles and responsibilities

2. Hadoop Architecture and Cluster setup 

Hadoop server roles and their usage

Hadoop installation and initial configuration

Deploying Hadoop in a pseudo-distributed mode

Deploying a multi-node Hadoop cluster

Installing Hadoop Clients

Understanding working of HDFS and resolving simulated problems.

3. Hadoop cluster Administration & Understanding MapReduce
 Understanding secondary name node

 Working with Hadoop distributed cluster

 Decommissioning or commissioning of nodes

 Understanding MapReduce

 Understanding schedulers and enabling them.

4. Backup, Recovery and Maintenance 
Common admin commands like Balancer

Trash, Import Check Point

Distcp, data backup and recovery

Enabling trash, namespace count quota or space quota, manual failover or metadata recovery.

5. Hadoop Cluster: Planning and Management
Planning the Hadoop cluster

Cluster sizing, hardware

Network and software considerations

Popular Hadoop distributions, workload and usage patterns.

6. Hadoop 2.0 and it’s features 
Limitations of Hadoop 1.x

Features of Hadoop 2.0

YARN framework, MRv2

Hadoop high availability and federation

Yarn ecosystem and Hadoop 2.0 Cluster setup.

7. Setting up Hadoop 2.X with High Availability and upgrading Hadoop
Configuring Hadoop 2 with high availability

Upgrading to Hadoop 2

Working with Sqoop

Understanding Oozie

Working with Hive

Working with Hbase.

8. Understanding Cloudera manager and cluster setup, Overview on Kerberos

Hive administration, HBase architecture

HBase setup, Hadoop/Hive/Hbase performance optimization

Cloudera manager and cluster setup

Pig setup and working with grunt

Why Kerberos and how it helps.

For whom Hadoop is?

IT folks who want to change their profile in a most demanding technology which is in demand by almost all clients in all domains because of below mentioned reasons-

•  Hadoop is open source (Cost saving / Cheaper)
•  Hadoop solves Big Data problem which is very difficult or impossible to solve using highly paid tools in market
•  It can process Distributed data and no need to store entire data in centralized storage as it is there with other tools.
•  Now a days there is job cut in market in so many existing tools and technologies because clients are moving towards a cheaper and efficient solution in market named HADOOP
•  There will be almost 4.4 million jobs in market on Hadoop by next year.

Please refer below mentioned links:

http://www.computerworld.com/article/2494662/business-intelligence/hadoop-will-be-in-most-advanced-analytics-products-by-2015–gartner-says.html