Hadoop is the next big thing when it comes to Relational DataBase Management System (RDBMS). Unlike other RDBMS like MS Access, MS SQL server, MYSQL, Oracle; Hadoop handle a large data size of Terabytes and Petabytes. While it is best for handling "BIG DATA" as it is called through Parallel Data Computations using the Hadoop Distributed File System and MapReduce Program; it is not best for handing lots of small data sizes.


Why Use Hadoop?
The essence of using Hadoop is summarised as the 4
characteristics - the 4Vs..

Volume: used for handling a VERY large Data (and still growing). The term "Big Data" is often attributed to it.

Variety: Capable of handling different data types. That is
- structured data
- semi-structured data
- unstructured data
NB: all other RDBMS handles ONLY structured data.
And 80% of the world's data is unstructured. Which leaves only Hadoop for manipulating 80% of the world's data while the other RDBMS handles just 20% only. (You can see it is a great invention)

Velocity: Computations is fast using Hadoop as it breaks down large data, send it across different nodes in the cluster and processes are done simultaneously. This method is called Parallel data computation.

Veracity: Hadoop is Reliable as it replicates data across the clusters in case of node or rack failure.





How old is Hadoop
About 5 years old!

It is a hot case profession that will be of high demand



Who uses Hadoop Currently?

-Google. - Facebook. - IBM. - Alibaba. - Yahoo. - Adobe. - Linkedin. - Ebay etc.




What is Hadoop used for?
-Big Data Exploration
-Data Warehouse Augmentation (now Data Warehouse Modernisation)
-Intelligence/Security
-Operation Analysis
-360^• view of a customer




When not to use Hadoop
-when work cannot be parallelized. That is there is inter-dependency of data.
-Not good for manipulating lots of small data sizes.
-Though it is not costly to setup, maintenance is.
-Low latency data access (since MapReduce is uesd to crunch data)
-Random Access Data (for transaction). On this, IBM's Adaptive MapReduce should be a solution to this (and prob. Low latency data access). I can't conclude yet (I am a student not expert) but I will be going through the insight.




Why this Thread?
Though, Hadoop is still new in the market, it brings hope to manipulating unstructured data. Hence, there is less "community areas" where the course is being discussed.

I hope this could be a talking point for Nairalanders who are students and Experts of Hadoop. We could learn together.


Front Page, Pls, Mods: seun, fayimora, javanian

Learn more visiting the [url=hadoop.apache.org/]official site[/url] or brief essence of hadoop.

Of course, you can ask google, lot of answers for you!




Those with the knowledge, please do signify and contribute. Let's share knowledge.

There are many Hadoop flavours but a few standout.
a. Cloudera
b. Hortonworks.

Hadoop itself is an umbrella project for many other projects. It consist of the filesystem(HDFS),query languages(Hive,etc),the core which is the hadoop framework which allows you to write map reduce applications etc.
The benefits of Hadoop is that you can cluster machines and use them for solving time consuming jobs.

@seunthomas. Especially a very BIG amount of unstructured DATA.
Are you currently a student? Or a profeSsional?

tohero:
@seunthomas. Especially a very BIG amount of unstructured DATA.

Are you currently a student? Or a profeSsional?

Professional

seunthomas:

Professional

If I may ask, a DATA SPECIALIST or an HADOOP EXPERT?

If I may, again, institute please? as majority of ict firms here are not yet acquinted wit the knowledge.

Where are others?

Welcome youngzedd, I guess you are a student of Hadoop.
"Others" are yet to showcase themselves, maybe with time they will.
Are you also a professional we can learn from? Probably we can start sharing our knowledge of the technology.

tohero:
Welcome youngzedd, I guess you are a student of Hadoop.

"Others" are yet to showcase themselves, maybe with time they will.

Are you also a professional we can learn from?
Probably we can start sharing our knowledge of the technology.

I want to learn it, just looking for materials.

Youngzedd:



I want to learn it, just looking for materials.

PDFs are available but Virtual Machine used for practicals are not; we were made to connect to IBM's server.

However Big Data University provide software you can use for practicals on courses of Hadoop which they provide. softwares are above 3GB and they can be run on Linux(x86-64) and Windows(64).

Hey Guys, I have worked on projects using spark, hadoop, storm and I have also used no-sql databases like cassandra. I am happy to help at anytime.

Though i'm still a student in this area. I belive that Hadoop is the future of high scale RDBMS