CA Technologies has adjusted its menu to focus on cloud, Agile and open source technologies on distributed platforms, but the company can’t forget what still puts food on its table.
At its annual CA World conference last week in Las Vegas, the company debuted a handful of modernized mainframe software tools with a healthy dose of artificial intelligence and machine learning technologies, which are intended to beef up security, analytics and automation and better manage corporate IT shops’ data.
“We hear mainframe users say, ‘We like this idea of driving more automation or speed and quality of product deployments,’ and so they want these capabilities on their platform,” said Stephen Elliot, analyst at IDC. “CA has to build this sort of foundation [with mainframe users] to maintain relevance with them over the long term.”
The company’s Mainframe Operational Intelligence incorporates machine learning and automation capabilities to capture patterns and trigger dynamic remediation. The product better enables IT shops to predict potential problems sooner and to automate a fix before service-level agreements are affected.
Other tools, Trusted Access Manager for Z and Dynamic Capacity Intelligence, use analytics and machine learning to restrict and monitor all activities by privileged identities on the mainframe. Technical support people who lack authorized credentials can still access sensitive data on a mainframe for a short period of time to fix a problem, such as an off-hours fix to payroll or sales systems.
These mainframe software tools, combined with the IBM System Management Facilities Adapter, allow IT shops to access data directly from IBM’s Z series of mainframes, including data that resides in non-CA applications.
“We are tying intelligent automation to performance and then combining that with human knowledge,” said Ashok Reddy, general manager overseeing CA’s mainframe business. This introduces flexibility to cure IT problems like a doctor, to decide the best prescription to fix a particular problem, he said.
An infusion of AI and machine learning into legacy mainframe software tools is a step in the right direction to give the venerable platform more street cred among IT shops that increasingly gravitate to DevOps and cloud platforms, according to some analysts.
“Getting mainframes to perform in a way that optimizes the business has always been difficult,” said Judith Hurwitz, president and CEO of Hurwitz & Associates in Needham, Mass. “But if they can work more like a transactional server supporting clouds and a variety of workloads by adding AI, then they can learn from those new patterns to make changes based on how they used to work and how it works now.”
Eventually, businesses and their applications will be driven entirely by business outcomes, and the underlying architecture won’t matter. “There will be certain workloads where the lowest-cost option might be the best architecture, [and] other workloads are so critical to the business that they should stay on the mainframe,” Elliot said.
And as IT shops increasingly connect their internal data to a wide variety of external data sources, through technologies such as containers and microservices, the threat of cyberattacks also rises.
Stephen Elliotanalyst, IDC
“We’re moving toward a world of services where everything is an API, and you can connect anything to everything,” Hurwitz said. “This issue of identity and trust and who can access what is something IT people have to worry about.”
And as IT shops increasingly gravitate toward multi-cloud strategies, they must invest in the management and coordination of clouds of every stripe and color.
“The reality is [IT executives] have to invest across all the swim lanes no matter if its private, hybrid or public clouds. They have to plan for management and orchestration across all the platforms,” Elliot said.
Mainframe software tools still represent the majority of CA’s overall revenues. But as mainframe IT professionals embrace modern technologies, the company has been forced to pivot its business to embrace cloud- and DevOps-based technologies. However, the company has more work to do to further crystalize its vision, Elliot said.
“There are some interesting ideas percolating there, but the question remains how does CA monetize it and bring it to market in a scalable fashion,” he said. “The next step — and it is a long time coming — is to take these interesting seeds and turn them into oak trees.”
Ed Scannell is a senior executive editor with TechTarget. Contact him at email@example.com.