Some consider the IBM Mainframe to be the last bastion of proprietary computing platforms, for obvious reasons, namely the CPU server architecture and the single manufacturer, IBM. The historical and legacy ability of said IBM Mainframe to transform Data Processing into Information Technology and still participating in the Digital Era is without doubt. However, for many, the complicated and perceived ultra-expensive world of software pricing generate concern, largely based upon Fear, Uncertainty and Doubt (FUD), which might have generated years if not decades of under investment for those organizations with an IBM Mainframe.
Having worked with the IBM Mainframe for 35+ years, I have gained a knowledge that allows cost optimization and contemporaneous usability, which given the importance of the IBM Mainframe platform to IBM from a revenue viewpoint, will safeguard that the IBM Mainframe will have a long future. However, the last decade or so has seen a rapid evolution in Open Source, DevOps, Enterprise Class Support for Distributed Platforms, Mobile and Cloud computing, et al, potentially generating an opportunity for the global IBM Mainframe user base to once again consider the platforms value proposition…
Let’s consider this server platform choice from a business viewpoint. On the one hand, there are the well versed market statements, where 80%+ of corporate data resides or originates from IBM Mainframes, while IBM Mainframes enable 70%+ of global commercial transactions, et al. In recent times there are global businesses, leveraging from the cloud or Linux Open Source technologies, to run their business. For instance, Netflix reportedly runs its media on demand business via the Amazon Web Services (AWS) cloud, while said platform is facilitating a Data Centre reduction of 34 to 4 for General Electric (GE). There are many other such “early adopters” of this commodity infrastructure provision opportunity, including Capital One, Hertz and Juniper, naming but a few.
Quite simply, the power of Mobile processors, primarily ARM and supporting software ecosystem empower each and every potential consumer with a palm sized smart computing platform, while the power and supporting software ecosystem of x86 processors, generate an environment for each and every global business, mature or not even launched, to deliver an eminently usable and scalable IT Infrastructure for their business model.
Of course, the IBM Mainframe can do this, it always has been at the forefront of IT architectures and always will be, but for the “naysayers”, its perceived high acquisition and running costs are always an easy target. As somebody much cleverer than I once said, timing is everything, and we’re now encountering a “golden sunset” for those Mainframe Baby Boomers, just like myself, that will retire in the next decade or so. Recently I was talking with a large IBM Mainframe customer, who stated “we’re going to lose 1500 years of IBM Mainframe experience in the next 10 years, how can you replace that resource easily”? Let’s just think about that metric; ~50 people with an average of ~30 years’ experience, but of course, they will all retire in a short time frame! You must draw your own conclusions as to that conundrum, how do you replace that level of experience?
In conclusion, no matter what IBM deliver from an IBM system z viewpoint, there is no substitute for experience and skill and no company, especially IBM has an answer to skills provision. In the last 10-20 years, Outsourcing or Managed Services has provided an alternative approach for some companies, but even this option has finite resource. If we consider the CFO viewpoint, where the bottom line is the only true financial metric, it’s easy to envisage a situation where many companies consider an alternative to the IBM Mainframe platform, both from a cost and viability viewpoint. As a lifelong IBM Mainframe champion and as previously stated, there will always be a solution for safeguarding the longevity and viability of the IBM mainframe for any Medium to Large sized business. However, now is the time to act, embrace the new Open Source, DevOps and Hybrid Cloud opportunities, to transition from a Baby Boomer to Millennial Mainframe workforce!
Is there an alternative approach and what is the Software Defined Mainframe (SDM)?
Put simply, SDM is a technology from LzLabs enabling the migration of mission-critical workloads from legacy IBM Mainframe environments to x86 Linux platforms. Put another way, LzLabs have developed a managed software container that provides enterprises with a viable way to lift and shift applications from IBM Mainframes into Red Hat Linux or Cloud environments. From my first glance, the primary keyword here is container; there was a time where the term container might have been foreign to the System z Mainframe, but with LinuxONE and zVM, Docker and KVM are now commonplace and accepted functions. The primary considerations for any platform migration would include:
- Seamless Migration: The LzLabs Software Defined Mainframe (SDM) ensures the key capabilities of screen handling, transaction management, recovery and concurrency are preserved without changes to the applications. LzOnline is capable of processing thousands of online customer transactions per second using commercial off-the-shelf hardware.
- Major Subsystem Compatibility: The LzLabs Software Defined Mainframe (SDM) safeguards 100% compatibility with existing job control syntax, and also enables job submission via network connected nodes that support conventional job entry protocols. LzBatch provides a full spool capability that enables output to be managed and routed in familiar ways. Use of conventional job submission models, with standard job control, also means existing batch scheduling can operate with minimal changes. Other solutions include LzRelational for Relational Database Management System (RDBMS) support and LzSecure, an authentication and authorization subsystem using security rules migrated from the incumbent IBM Mainframe platform.
- Application Code Stability: An innovative approach that avoids the requirement to recompile or rewrite legacy COBOL or PLI application source code. Leveraging from functionality delivered by Cobol-IT and Eranea, a simple and straightforward process to convert and potentially modernize existing application source code to Java.
The realm of possibility exists and there are likely to be a number of existing IBM Mainframe users that find themselves with challenges, whether retiring workforce or back level application code based. The Software Defined Mainframe (SDM) solution provides them with a potential option of simplifying a transition process, with seemingly minimal risk, while eradicating any significant dependence on another Distributed Systems platform supplier, during the arduous application source and data migration process.
From my viewpoint, I hope that this innovative LzLabs approach is a wake-up call for IBM themselves, who continue to deliver a strategic Enterprise Class System z platform, with all of its long term challenges, primarily cost based and the intricate and over complicated sub-capacity software pricing structure. Without doubt, any new workload can easily be accommodated for low cost via the recent LinuxONE offering, but somewhere along the line, IBM perhaps overlooked a number of Small to Medium sized customers, who once might have used entry level or plug-compatible platforms, including and not limited to S/390 Integrated Server, MP3000, FLEX-ES zFrame, T3 Liberty, et al. Equally from a dispassionate viewpoint, I welcome the competition of the LzLabs Software Defined Mainframe (SDM) offering and I would encourage all CIO and indeed other CxO personnel to consider the merits of this solution.