The web has dozens of prediction blogs and articles covering every aspect of technology, but although they’re interesting, how often do you go back to see whether those predictions were right or wrong? In the spirit of continuous learning, we looked at those four predictions about mainframes we made a year ago.
Prediction #1: Mainframe DevOps Will Become Mainframe DevSecOps
In 2017, we predicted that mainframe DevOps would continue to shift left and pick up security to include DevSecOps. We were right, though we didn’t go far enough. What we’re seeing is that in addition to DevSecOps, our industry is rapidly going from the concept of siloed mainframe-only DevOps to a cross-enterprise DevSecOps.
Companies clearly recognize the importance of security in the DevOps process. On the basis of the latest IBM research, 67% of mainframe users concur that mainframes are now more connected to the outside world and are therefore more vulnerable. Additionally, 80% say the growing sophistication of cybercriminals is a major reason to increase mainframe security.
Vulnerabilities abound, which is why identifying sensitive data on the mainframe remains highly critical: information leakage is the most prevalent vulnerability, with 66% of all apps failing on the first scan.
Switching from DevOps to DevSecOps is a major step toward resolving these security concerns. DevSecOps means applying application security best practices earlier in the development life cycle by including security and compliance experts from the beginning. The cross-enterprise aspect of DevSecOps has emerged rapidly because mainframe back-end applications are being closely integrated with web and mobile applications that support various lines of business (LOBs). The LOBs drive frequent changes and upgrades to the applications they use, demanding greater mainframe agility and security.
There’s another reason for the speed at which cross-enterprise siloes are being eliminated. As seasoned mainframe developers are set to retire and the next generation comes up the learning curve, the need to provide a common set of cross-enterprise tools becomes more critical. And that’s what we’re seeing from customers, be it in the area of DevSecOps or in the area of entirely new service catalogs that enable self-service using popular open-source tools for new developers.
Prediction #2: Mainframe as a Service (Mainframe Cloud) Goes Mainstream
We expected mainframe as a service (mainframe cloud)—which provides compute, storage and DevOps capabilities as self-service—to flourish in 2017. Although it has yet to become mainstream, we’re seeing greater use of this approach by early-adopter customers as they seek to address skills gaps, improve economics, increase agility and more. Increasingly, our customers are seeking the flexibility to consume mainframe as a service rather than deploy single-tenant, on-premises or operational models.
We anticipate continued growth in this area as the following takes place:
- The business imperative for cross-enterprise solutions with greater time to market drives demand for application modernization, software-delivery automation and enterprise DevOps transformation.
- The ability to optimize operational costs becomes a competitive differentiator for businesses.
- The need to train a new generation of developers and engineers demands a cloud-like user experience that allows users to employ their preferred tools from their desktop or mobile device.
- Buyers increasingly want flexibility and transparency, such as they’re familiar with in open systems.
Put simply, business agility remains the critical factor in employing the mainframe to shape next-generation business applications.
Prediction #3: Mainframe Becomes the Hub for Automated Intelligent Systems
Gaps and inefficiencies in IT operations continued to plague businesses in 2017, with enterprises losing an average of $21.8 million per year in downtime and with the average cost of downtime ranging from $140,000 per hour to a whopping $2.5 million per hour.
Even though the mainframe remains reliable, companies are still looking to increase efficiencies across their skills and to boost speed. That’s why our prediction has proven correct, and we’re finding that augmented intelligent automation is delivering on a vision of a self-driving data center: in production, not just in theory. “Algorithmic IT operations (AIOps) platforms,” a term coined by Gartner, are giving businesses the following:
- Abnormal pattern detection to predict problems earlier
- Faster problem isolation with insights from multiple data sources
- Pattern recognition to trigger automated remediation of “known” issues
- Operational feedback to guide the next best action
Early adopters of intelligent automation are achieving financial benefits while also delivering a more seamless customer experience. For example, customers are using mainframe operational intelligence to achieve 5x improvements in resolution time and mean time to resolution (MTTR).
Surprise Prediction: Blockchain Will Cause a Resurgence of Mainframe Interest
Our final prediction in 2017 addressed the anticipated ramifications of blockchain technology on the mainframe. And it did come true. Blockchain investment has caused a mainframe growth and innovation to resurge as enterprises seek to implement digital trust at speed and scale. Blockchain-driven applications will soon be one of the largest users of capacity from the 60+ data centers that IBM rents globally, and one-third of banks, retailers and health-care institutions are expected to use blockchain by 2021. Many companies are evidently taking that route, as IBM recently announced that IBM Z revenue is up 71% year over year and that it has shipped the most mips in history—and these results are just for the first full quarter since the launch of the Z14, with its pervasive encryption capabilities.
Buy why mainframe? You might already know that mainframes contain 71% of corporate data. You can probably guess the rest. Scalability is what makes the mainframe a major platform destination. Currently, blockchain implementations function at 5–10 transactions per second; the goal is to enable more than 1,000. A good solution is to build blockchain technology on the foundational support of the mainframe.
These benefits are large and transformational. For example, HSBC, a member of the Digital Trade Chain consortium, is testing a large blockchain as part of its initiative to restructure its cost base by digitizing everything and eliminating paper. The goal is to reduce trade costs by up to $2 trillion by providing a transparent information-sharing mechanism via blockchain.
2018 and Beyond
So, we weren’t too far off. The question is, where do you want the mainframe to take your enterprise in 2018 and beyond? I invite you to visit us at www.digitaltrust.ai to learn more or reach out to me directly.
 Veracode, State of Security report.
 Naegle, Robert, “IT Market Clock for IT Automation.” Gartner, September 2016.
 IDC FutureScape: Worldwide IT Industry 2018 Predictions.
 Source: IBM Earnings Report and Call 4Q17.
About the Author
Ashok Reddy is General Manager, CA Software Mainframe Business Unit. Ashok is responsible for CA’s entire portfolio of mainframe software products and WW product organization. He applies his 30 years of software-industry and technology expertise to the mainframe business, where he is setting the vision, strategy and direction to drive the organization to help CA’s customers solve their most difficult business problems in the application economy and in their journey of digital transformation.
Ashok joined CA in November, 2015, from IBM, where he served as the VP of Offering Management for the company’s API Economy and Hybrid Integration portfolio. He was at the forefront of leading many highly strategic development efforts for IBM, including API Economy, their first-ever SaaS offering; the Jazz ALM platform; DevOps for Mainframe; next-generation mainframe compilers; IBM Bluemix DevOps (PaaS); and IBM’s IOT and Mobile Development Solutions. Previously, Ashok worked at Rational Software, Honeywell and Novartis.
Ashok has a MBA, MS and BS in engineering and is Professional Engineer (PE, California). He has patents in IT governance and automation and has published several papers.