How mainframes fit into hybrid environments
Cavan Images / Getty Images
Connecting state and local government leaders
Integrating cloud and on-prem mainframe platforms helps agencies deliver new services.
A mainstay of agencies’ information technology departments, mainframes not only have staying power in modern environments, but they are getting upgrades, too.
“Mainframe in general is still a massive part of government business,” said Susan Wedge, managing partner for the U.S. Public and Federal Market at IBM Consulting. “The number of states utilizing mainframe has continued to remain consistent, and so we’re seeing mainframe not as something of the past, but as a part of the IT modernization for our clients moving forward.”
“The conversation with every one of our clients, regardless of industry, is about cloud,” said Barry Baker, vice president of Product Management, IBM Z and LinuxOne. But customers realize that some workloads are well-suited for cloud and others, especially those with large volumes of data and high security needs, must run on premises. The result is clients opting for a hybrid environment with fit-for-purpose platforms.
The task at hand is driving integration among the platforms while delivering new services, he said.
One way to do that is through updates to the mainframe technology. For instance, IBM introduced last month the z16 mainframe, designed for hybrid cloud environments; it features an on-chip artificial intelligence accelerator and what the company calls the industry’s first quantum-safe system. That means people can’t harvest the data now and decrypt it later when quantum machines have that capability, Baker said.
IBM releases a new generation of the platform every 2 to 2-1/2 years, he said, and commercial customers are often faster to take advantage of the changes than government ones.
“That has its pluses and minuses,” Baker said. “In some cases, some of the first users end up learning and figuring things out. On the cloud point, many of the banks that we have worked with – five years ago they were trying to figure out how to move to the cloud. Now they are helping us carve out what the new world is. It’s more about an adoption cycle and governments being more constrained than the commercial sector.”
On the developer side, a challenge with mainframes is the belief that they don’t support the use of new tools, but that’s not the case, Baker said. Developers can use the same toolchain to develop for Amazon Web Services or write Java or COBOL for the mainframe.
“From a developer’s perspective, you want to use a common toolchain that you’re familiar with, so you want to use Git for a source code repository, you want to use things like VS code for doing actual software development as your [integrated developer environment],” he said. “You want to use Jenkins to have a completely automated [continuous iteration/continuous delivery] pipeline. All those things are possible,” he said. “We have clients that are doing them on the platform, which enables them to move at the same speed regardless of what platform they’re using.”
Additionally, the platform is built to handle data- and transaction-intensive workloads, and as customers create new engagement apps, they’re laying down a standard REST-based API over the top of the environment so that new mobile apps can call and use those services quickly.
For example, State Farm, an insurance company, wanted to speed the development of digital services that integrate with core systems. To do this, the company standardized development cycles across all enterprise systems using “open source, home grown and proprietary tools to provide a modern integrated DevOps system for maintaining existing core applications and building new functionality,” according to a case study.
On the data side, commercial clients are seeing benefits in colocation and consolidation of some workloads that may run on other platforms but map well to the mainframe, Baker said, citing the ability to consolidate 20 x86 servers onto one on the IBM Z platform.
“It’s not cloud or mainframe, it’s cloud and mainframe,” Wedge said. “It really is what applications make sense in which environment because even in the cloud it’s what’s in the public cloud, what’s in the private cloud. So I think it’s an and vs. an or, and it really goes to that different workloads being optimized in different environments to be able to deliver the value that they need to deliver.”
Stephanie Kanowitz is a freelance writer based in northern Virginia.