Mainframes: Should You Stay or Should You Go?
It's a more difficult question than you may think in the current IT world.
- By Dan Kusnetzky
A short while ago, I wrote an article discussing a recent BMC-sponsored study that focused upon mainframe use within enterprises. This study came to the conclusion that mainframe use is likely to continue, and even increase somewhat. I had no idea that this article would cause a number of suppliers, including Compuware, COBOL-IT and TmaxSoft to reach and present their views on the subject.
Compuware's viewpoint was that enterprises can continue to rely on their mainframes and even increase their reliance on these systems through the use of DevOps practices and Compuware's toolset.
COBOL-IT and TmaxSoft took the other side, and suggested that with their tools it's a straightforward process to move enterprise COBOL applications to other platforms, such as Windows, Linux and midrange UNIX offerings. They support converting mainframe COBOL to either COBOL on the target platform or to JAVA or .NET as a platform.
For the most part, however, these folks never mention that fact that many mainframe workloads include COBOL, PL/1 and, possibly, mainframe assembler. It's even possible that other languages, such as RPG, FORTRAN and a few others might be in the mix as well. For the most part, these suppliers don't explain what enterprises are to do with these applications or portions of a given workload.
Suppliers of midrange UNIX and industry-standard x86 platforms have long targeted mainframe installations with stories of how a migration from the mainframe to their platform will improve performance, lower costs and decrease levels of complexity.
Are Mainframes 'Legacy'?
These suppliers have even coined a marketing buzzword -- "legacy" -- to describe mainframes. It's as if the use of a word would discredit mainframes in the face cost of ownership (COO) and return on investment (ROI) studies that show that, in certain circumstances, that mainframe computing environment may actually be less costly and complex than moving workloads from a single mainframe system to a herd of industry-standard computers.
Enterprise, however, have kept their mainframes; and, if the BMC study is to be believed, are increasing their use of the platform. "Why are they staying on the mainframe?" is a legitimate question in the face of all of the market activity to get enterprises to change.
As I've pointed out in articles, such as Are Enterprises Running Away From The Mainframe? BMC Survey Says 'No', Walking Away From the Mainframe, and DevOps and Mainframes Don't Have to be Enemies, migration decisions are complex and, on many occasions, lead to a computing environment that didn't meet the goals of cost savings or reduction in complexity.
Dan's Take: The Big Mainframe Questions to Ask
Enterprises have to consider if, in their IT environment and with their applications, that a distributed computing environment would be better than a centralized one. They also must consider if a patchwork quilt management and monitoring environment will introduce simplicity and cost savings over a more centralized management approach. They should also consider security and reliability issues that might arise in a distributed, multi-platform, multi-vendor computing environment.
Other suppliers are beginning to reach out to me to tell their stories, and I expect to learn about things such as mainframe to cloud (which in all likelihood includes a platform change as well), and migration of mainframe applications to other development platforms.
Is your organization considering a mainframe-to-something-else migration? If so, what platforms are likely to be the target?
Daniel Kusnetzky, a reformed software engineer and product manager, founded Kusnetzky Group LLC in 2006. He's literally written the book on virtualization and often comments on cloud computing, mobility and systems software. He has been a business unit manager at a hardware company and head of corporate marketing and strategy at a software company.