Summary
The
core intention of writing this blog is to describe features of legacy mainframe
and the emerging era of Big Data. It also demonstrates the process of
offloading data to Hadoop.
Advantages
of Mainframe:
Legacy
systems used by many organizations are more secure, scalable and reliable
machines which are capable of tackling huge workloads.
Mainframe
handles even mission-critical applications with minimal resistance, like
processing banking transactions, where both security and reliability and
security are equally important.
Drawbacks
of Mainframe:
They
sustain Large hardware and software.
Processing
Prices.
Many
organizations, in the current era, take the urge to initiate a part of
migration and continue the same in all the aspects of business applications to
newer reliable platforms.
This
process helps organizations to reduce costs incurred and meeting the evolving
needs from the business
Advantages
of Big Data technology over Mainframe:
Cost
Effective.
Scalable.
Fault
Tolerant.
The
cost in maintaining and to process the mainframe can further be reduces by the assimilating
a layer of Hadoop or to completely off load the batch process to Hadoop
Similarities
of mainframe and Hadoop are as below :
Distributed
Frameworks
Handle
massive volumes of data
Batch
workloading
strong
sorting
Business
Benefits of adopting the Big Data Technologies along with Mainframe or over
Mainframe
Using
Hadoop ecosystems like PIG , HIVE or MapReduce the Batch processing can easily
be done.
Jobs
from the Mainframe systems can be taken and processed on the Hadoop the output
of the same can be viewed at the mainframe end reducing million instructions
per second (MIPS) costs.
(MIPS
is a way to measure the cost of computing: the more MIPS delivered for the
money)
Organizations
look at return on investments at every corner during up-gradation or migration.
similarly Migrating mainframe to Hadoop is this condition met due to minimal
infrastructure , the batch process costs and flexible upgrade of applications.
Process
of Offloading Data to Hadoop
Offloading
approach is recommended in the following simple steps.
To
create Active Archives and copies of limited mainframe datasets in the Hadoop
distributed File system.
Secondly
to migrate larger amount of data from the source from sources like
semistructured data sources or Relational DBs
Final
iteration of moving the expensive mainframe batch workloads to the
much-sophisticated Hadoop