MapReduce and Design Patterns - Input Output Pattern Overview

In this video, we are going to discuss input-output pattern overview

So, in this particular pattern, we shall be discussing how the MR tasks can store their data

on the SDFS or onto some external storage for the future read and write operations

Now, what is input-output pattern? Sometimes we do not use the traditional method to store our data onto the SDFS

so the MR jobs cannot read the data block from the Hadoop file system next time because it is not getting stored onto the SDFS so obviously in future it is quite impossible to get those data from the SDFS directly so like this situation we need to organize our MR job to perform read and write operations from some different storage sources so in this way you can find that either the MR jobs can store the data on the Sdhapes

for the future retrieval but if they are not storing the data on the sdFS then obviously the

data accessing will be a problem in the future days so that's why it must be capable to

store the data onto some external storage spaces so in this section we will see how to randomly

generate data and store it onto the sdfs so we'll be looking for this how to store our data

randomly generated data onto the sdfs and also we shall learn to use

external source to read and write using them so this is our input output design

pattern we'll be discussing lots on this particular topic in the next successive

videos please watch all of them and thanks for watching this