Modernization Carries Varied Meanings for Different People
Like anything else, modernization depends on context and perspective.
By Jim Zell03/23/2020
Defining ModernizationEvery CIO who has even the smallest mainframe footprint would see this device as an asset as a precious tool that must run all critical workloads. Right? Getting one or more of the fastest, most resilient and highly available processors would mean you were operating on the latest and greatest technology available. What I see however, is modernization carries different meanings for different people. Like anything else, modernization depends on context and perspective. Ask 10 people about a particular topic and you are likely to get 10 different viewpoints. IT modernization is no different. To many, modernization is simply moving everything to someone else’s hardware, the cloud. To others, modernization refers to the latest gadget.
Modernization and IBM Z CustomersWhen I have discussions with IBM Z customers who work in the trenches and strive to harness the power of the mainframe, they think about increasing application availability to improve the end user experience. They want higher availability and resiliency so they can reduce or eliminate the mean time to repair (MTTR). They want to shorten mainframe IPL windows. This thinking leads them to specific mainframe modernization efforts such as:
- Sysplex-enabling infrastructure components like CICSplex, IMSplex and Db2 Data Sharing
- Speeding up IPLS by utilizing one-touch IPL or automating the manual tasks associated with IPLs
- Making sure they’re fully utilizing coupling facilities
- Squeaking MIPS/MSUs out of their system by scrutinizing SMF data
- Ensuring they have redundant hardware, started tasks that can run on multiple LPARs, LPARs that can run on multiple CECs, etc.
- Implementing the latest security and cryptographic systems and protocols
Modernization From an Analytics PerspectiveModernization to others is completely different. Other folks may not even know what a sysplex is or care. That is not a problem or wrong—it’s just a different perspective. This group thinks of harnessing the data that is buried in those mainframe system of record files and mining it for information that can be used to improve business for customers. These are the analytics folks. They think about AI and machine learning as often as some of us think about sysplexes. To them, modernization is simply about accessing data in new and creative ways to harness the power it can provide to their business. Often, these folks aren’t even aware that that the analytic prowess they strive for is available on the mainframe. The data need not be moved anywhere, like to the cloud to perform analytics on it. So, these folks have a different view, but it is still a modernization view.
Modernization for Midrange DevelopersThe next group is midrange developers. Many of these experts see the mainframe as old software that nobody understands how to program. These ideas are incorrect, but they’re really just thoughts of the uninformed. To this group, modern means one thing: modern tooling to do a job. Green screens are anathema to this group. It doesn’t matter how much TSO has been upgraded, how resilient it is or how many instructions have been added to the mainframe that COBOL can now utilize to improve performance by 40%. They’re not focused in architectural improvements in mainframe chip design. Rather, they’re focused on whether or not the machine can be accessed with tooling that has built-in multiple panels for different coding languages, access to Git, access to other environments, debug capability, etc.
Cloud ModernizationThe last group thinks that modernization means cloud. There’s a movement to move mainframe workloads to the cloud—that is, to take them off the mainframe and put them on x86 machines. Companies and consultants will promise to provide modern eclipse-based tooling once workloads are moved to the “modern cloud.” They will even convert some of your code to Java so it can run on an x86 machine. The software that won’t convert—well, you can keep your COBOL and still run it on an x86.
The rub is this: all of those promises can be fulfilled with the data and programs staying where they are, on the mainframe. You can do analytics, machine learning, AI, use eclipse-based tooling (an IDE!), run Java (and a half dozen other languages) all on the mainframe.
The Mainframe Is a Modern Access, Modern Tooling OptionThe mainframe requires experts to maintain and manage it. Today, however, that machine marvel can be made available to every developer through Eclipse-based tooling. The same developers can still use their modern languages. Your system of record data doesn’t have to move, sucking up network bandwidth and becoming more out of date with every transfer. You can perform analytics on that data in place, right now.
Whether it is modern access, modern tooling, the popular programming languages or the latest AI processes, the mainframe can do it all, today. Think about it. If you can do everything you want with the device and data where it is right now, why go to the expense of moving it somewhere else?
Jim is an IBM Chief Engineer and executive architect. He leads mainframe architecture in the public sector for IBM GTS customers in North America.
Sponsored ContentAchieve Compliance Without Impacting Productivity
Post a Comment
Note: Comments are moderated and will not appear until approvedcomments powered by Disqus