This blog is for Mainframe Modernization professionals to discuss and share perspectives, point of views and best practices around key trending topics. Discover the ART and science of mainframe modernization.

« Transform the IBM i-series/AS400 with DevOps: It's possible | Main | Why is Legacy Modernization a key Influencer for your Tech. Strategy this year? »

Big Iron for Analytics

blogfinal.jpg
Some like to call it the 'Big Iron', while a few refer to it as a dinosaur- a callback to its invention in the 1950s. The Mainframe has come a long way. Let us take some market trends and how mainframe copes up with them.  

What are the current trends driving the market?
  • Data: Is growing exponentially, with newer data types
  • Data Generators: There are newer streams through which data   is being generated like smart devices, GPS location of users,  logs etc. 



Technology trends
  •        Nearly half of total sales is influenced by Digital organizations
  •        Nearly 20% of organizations that go digital tend to have an upper hand over competitors

Data Insights
  •        75% of CIOs will recognize the limitations of traditional IT and embrace a leadership approach that embodies a          virtuous cycle of innovation
  •        40% of IT projects will create new digital services and revenue streams that monetize data
  •        zAnalytics is available from model Z13 and upwards of machines
  •        For real time analytics, additional IBM products for quick Db2 table access are needed
  •        Will have to find a way to store newer data types in Mainframe


Is the Big Iron capable of addressing these trends and associated requirements?

The answer is "partially yes". Depending on the specific use case, a few will be suitable for, and addressed through zAanalytics.  

But first, what is zAnalytics?

It is an offering from IBM, that enables organizations to undertake analytics and machine learning on the mainframe. 

Advantages of zAnalytics

  •       Data In-place: There is no need to move data out of mainframes as zAnalytics
                        - Is capable of reading data from DB2 table, VSAM, PS files , IMS DB
                        - Avoids data duplication and latency
  •       Security: Users continue to benefit from the highest level of security that Mainframes offer
  •       Engine:  It uses the Apache Spark engine to divide and process the data in-memory
  •       On-line and batch: Both online and batch workloads are supported for analytics
  •       Specialty engine offloading: Few of the workloads can be offloaded to the zIIP engine, which further reduces the         cost of execution

ยท         What are the constraints of zAnalytics?

  •          zAnalytics is available from model Z13 and upwards of machines
  •          For real time analytics, additional IBM products for quick Db2 table access are needed
  •          Will have to find a way to store newer data types in Mainframe 

To evaluate zAnalytics, we have written a simple application with a simple use case. 

Let's explore the use case we are looking at:

It's a simple application as follows

  1.            Customer File: This file consists of 1 Million records consisting of information below
  2.             Transaction file: A file containing 10MM transactions of the customer
  3.             Use case:        
      1. Load both data file into data frames
      2. Join Both data frames on Unique customer ID
      3. Sum the transaction amount at customer ID level
      4. Display top 20 users who spent the maximum                                                    

The hardware used for this test was a zTrail machine provided by IBM. However, it is not meant for benchmarking.

zAnalytics was able to perform this workload at a rapid pace, in close to three minutes and thirty seconds (3:30).  

Conclusion:

Rewrite your workloads in zAnalytics and witness significant reduction in execution time and cost.

For more details contact: modernization@infosys.com


References:

[1] http://www.gartner.com/imagesrv/symposium/barcelona/docs/SYM_2012_Barcelona_BIIM_Summary.pdf

[2] https://www.idc.com/getdoc.jsp?containerId=US42256117

[3] https://www.idc.com/getdoc.jsp?containerId=US42256117


Blog authored by

Ramanath Nayak, Lead Consultant, Legacy Modernization

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)

Please key in the two words you see in the box to validate your identity as an authentic user and reduce spam.