Subscribe

Take analytics to the data, not vice versa

By Tracy Burrows, ITWeb contributor.
Johannesburg, 16 Nov 2018
Karmjeet Kahlon, VP, Worldwide z Hybrid Cloud, IBM.
Karmjeet Kahlon, VP, Worldwide z Hybrid Cloud, IBM.

Traditional approaches to data management and analysis, whereby data is exported for processing and analysis, are out-dated, over-complicated and costly, says IBM.

Speaking at the IBM "Unleash the Heart of your Enterprise" executive forum in Johannesburg this week, IBM experts said it is more efficient to process and analyse data where it is created.

Said Karmjeet Kahlon, VP, Worldwide z Hybrid Cloud, IBM: "We say: take the analytics to the data, don't take the data to the analytics. Copying and extracting data allows it to lose value and context, and causes latency, which makes the data out of date."

In addition, extracting data multiple times for various processes and analytics is inefficient and puts the data at risk.

"The average set of data moved off the mainframe gets copied 11 times," said Tom Ramey, director, z Analytics at IBM. He noted this added complexity, latency and risk to the entire environment.

"With 11 copies of the data, you lose control of that data and create a huge security risk." Moving vast quantities of data for processing and analysis also adds significant cost, he said.

Kahlon noted enterprises are advancing their use of data to deliver business value. "Data is the new oil, and enterprises have tons of it. They need to find new revenue streams out of their data. If they don't, they will go out of business."

At the same time, most enterprise data resides in mainframes. Martin Blignaut, enterprise and mainframe software sales leader, IBM South Africa, noted: "The mainframe environment is in place in major enterprises and public sector organisations. It's a system that actually runs the entire country and the economy. Mainframes are incredibly secure and can scale incredibly large."

Around 80% of the world's corporate data resides on mainframes and 91% of CIOs say their new customer-facing applications are also accessing the mainframe, IBM said.

Mainframe modernisation that allows for virtualisation, machine learning and artificial intelligence to support analytics within the database environment could significantly speed up data processing, reduce costs, and position organisations to innovate using their data.

IBM outlined its zSystem and enterprise analytics solutions designed to provide real-time insights from enterprise data into business opportunities, including:

IBM Db2 for z/OS: An enterprise data server for business-critical transactions and analytics.

IBM Db2 tools for z/OS: Utilities and tools to maximise database performance and availability, reduce cost and simplify data management.

IBM Db2 Analytics Accelerator: A workload-optimised appliance to integrate analytic insights into operational processes.

IBM Machine Learning for z/OS: A machine-learning solution that infuses continuous intelligence throughout the enterprise.

IBM QMF: A zero-footprint, mobile-enabled and secure business-analytics solution with visual reports and dashboards.

IBM Data Virtualisation for z/OS: A data management solution that provides applications with the ability to access, update and join mainframe data with other enterprise data in real-time, virtually, using modern APIs.

Share