Mainframes are powerful and sometimes underappreciated assets in today's enterprise architecture as software architects focus on data centers powered by x86 machines. But the fact remains that most of the corporate data still lives on the mainframe. The mainframe is still a computing powerhouse. A single installation can process 2.5 billion transactions daily, which equals to some 100 Cyber Mondays.
But together with power come challenges that need to be addressed by companies looking to modernize their mainframes. Every modernization process carries some risk, and that's why organizations should always turn to an experienced team to take care of it.
But before we go into that, let's take a closer look at the challenges that mainframe modernization poses to engineers.
1. Thinking that the technology alone will solve the problem
Mainframes are powerful; that's without question. But the applications running on them are only going to work because the development lifecycle process is well developed as well.
Some mainframe providers adopt approaches in line with agile and continuously deliver improved software that meets user needs.
But addressing the real needs of customers can be a challenge. Sometimes mainframe modernization might miss the mark. Even if your intentions are good, customer misalignment can still happen. For example, if stakeholders are very enthusiastic about a potential technology, they start focusing on it instead of what it would mean for the customer's needs.
One way to deal with that is to produce customer documentation that actually helps customers learn what they need to know instead of learning everything there is to know about the solution.
Take the time to understand the pain points and needs of your clients. Bring in the technology and innovation resources that actually address these problems. Meeting the real needs of the end-users is your goal in developing any technology, including mainframes.
2. Integrating database administration into modern DevOps practices
In the past, a database developer would often change the name of the column in the table only to have that one change introduce chaos to downstream SQL statements using that table. Most of the time, there was no direct connection between the code and database. Aligning database development with application development used to be very difficult.
But this changed with the emergence of DevOps and especially the mainframe landscape. One of the modernization challenges is developing a mainframe without broadening the team to include database administrators.
You can deploy agile development with database administrators being part of your teams alongside developers. Increased collaboration between the people who code the application and manage the database will have a positive impact on your modernization project.
With agile practices, DevOps increases efficiency in any software development lifecycle, including mainframe development. Including database administrators in the daily tasks of the developers is the easiest way to launch seamless change management in your mainframe development.
3. Identifying and reducing shadow IT projects
The main benefit of mainframes is the amazing computing capability. But for many companies, they are also large black boxes with logic and data embedded in them. IT departments might be struggling with this type of isolation internally, in particular the isolation around data.
Groups outside of mainframe development might address that problem by copying data out of the mainframe to other environments. This is called shadow IT. Shadow IT happens when you have a mainframe and a team ends up copying what the mainframe is doing because it wants independence and more control.
Usually, the data is copied to another target like an x86-powered cloud platform. The idea is to use this data for autonomous purposes.
It might seem that shadow IT isn't such a big problem. But it can generate enormous costs. For example, an organization might end up spending 200 million per year on the mainframe then spending another $100 million on distributed items that are actually based on taking data off the mainframe.
Instead of resorting to shadow IT, share the data and make the most of the available technologies. Encourage teams to share it, not copy it. Manage all the data together and share resources around resilience, transaction programs, and security. By launching this type of collaboration, you'll avoid placing your IT teams in silos – and potentially save millions of dollars.
4. Improving tool standardization
When integrating competing technologies, you might run into many problems, especially if you're facing cultural or technical disparities.
This problem is something we see happening to financial services institutions that have both legacy mainframes and emerging x86-based technologies that need to be supported. The IT budgets may be running into hundreds of millions, if not billions of dollars.
It's very likely that there are different products with different development histories and a conflicting architectural setup in your organization. The team might start using open-source tools like Jenkins, Eclipse, or Git. These tools have some duplicity with closed-source tools and might bring more architectural conflict. Integrating all of these products and technologies might become even more difficult because of automation that is now entering the world of CI/CD processes.
Automating the process means that it needs to be formalized and standardized. It's necessary to know how all the products in that flow will be interacting with each other. If you don't know that, integration is going to become the primary bottleneck in any of your automation efforts.
Lack of standardization also slows down the agility of a mainframe. Businesses that are looking for faster deployment choose x86 platforms that adhere to general conventions based on open-source standards. This is one of the biggest inhibitors to the mainframe becoming as mainstream as the cloud. The lack of tooling that delivers a great agile experience for mainframes is hard to find since most of the tooling is still proprietary.
5. Avoiding poorly-directed modernization efforts
Modernization essentially comes in two types. The first is when brand-new technologies are used to address the real shortcomings in the company's IT portfolio. The second type is when the new technologies just rehash the capabilities that have already existed in legacy systems.
Addressing the shortcomings in a cost-effective way is a smart move. Applying changes to a technology stack for the sake of it rarely comes with benefits. And yet, it happens often among organizations with mainframe computers.
Mainframe modernization may fail when you bring in modern tools and languages to the mainframe without knowing what's happening there already. This type of redundant development activities is a problem across many IT departments. Most of the time, it arises due to poor communication. That's why communication must happen beyond emails and weekly meetings.
Sharing information and using the wisdom of the crowd is critical for ensuring that the modernization efforts aren't misguided and you're actually investing in a mainframe that will support your business mission.
What are the risks involved in mainframe modernization?
COBOL is a programming language at the core of the mainframe industry. 95% of all ATM applications use it today. But COBOL is also the cause of its main problem. The growing skill gap is a real threat because COBOL experts are now leaving the workforce in greater numbers.
Education and training programs aren't producing a replacement for these developers anymore. As Baby Boomers retire, the hunt for COBOL developers is only going to intensify and rise into the main people risk.
Rapidly rising costs associated with the ongoing mainframe usage is another key risk organizations face today. In fact, mainframe maintenance costs are usually the fastest-growing component of an IT spend.
The annual software and infrastructure fees paid to mainframe vendors can approach billion-dollar levels for the largest users. Reducing these costs and other financial risks is the primary driver behind modernization efforts today.
Many mainframe applications still offer excellent performance, reliability, and security. However, they also might become an obstacle for businesses to adapt and innovate quickly.
Monolithic mainframe applications are difficult to change. Just consider the traditional and time-consuming Waterfall approach, interdependencies of large applications, or the daunting task of testing all the changes and enhancements.
Meanwhile, these companies are facing competition from cloud-native startups that are emerging in practically every industry and disrupting them with quick feature releases and modern capabilities used for delivering a fantastic user experience. Even industry leaders need to start addressing this problem.
It goes without saying that mainframe modernization comes with significant challenges. However, the reward part of the equation is rising together with the risk profile. In other words, investing effort in modernizing your mainframe is a smart move.
But you need expert advice to make it work. If you're looking for a reliable partner to help you out in the mainframe modernization effort, get in touch with us. At Maxima Consulting, we have years of experience in helping the financial services sector and beyond in modernizing mainframes in line with modern IT standards.