CIO.com Virtual Roundtable: Mainframe Application Modernization Initiatives in Support of Successful Transformation
Businesses that leverage their investments on the mainframe, while developing a modernization roadmap have an overall better outcome -- quantitatively and qualitatively -- than those that move off them.
Mainframes have been a workhorse in the financial services sector for decades and appear to be positioned to remain relevant for years to come.
The IBM co-hosts suggested that the best path forward for most financial institutions will likely revolve around the concept of “encapsulation” to develop mainframe modernization initiatives that start with a service interface.
Barry Baker, IBM
Businesses that leverage their investments on the mainframe, while developing a modernization roadmap have an overall better outcome -- quantitatively and qualitatively -- than those that move off them. These were the conclusions of a 2021 IDC study that analyzed 446 businesses across the globe to measure the efficacy of migrating to the cloud compared to modernizing existing investments in mainframe systems and applications. The case for the latter won the day. Re-platforming, the researchers found, can be riskier and more costly than modernizing existing investments.
In a recent CIO.com Chatham House Rule virtual roundtable co-hosted with Aparna Sharma and Barry Baker from IBM, technology leaders from around the country gathered to discuss how mainframe application modernization strategies are evolving to play a critical role in ensuring successful business transformation initiatives.
Here is what they had to say:
Mainframes have been a workhorse in the financial services sector for decades and appear to be positioned to remain relevant for years to come. As one participant noted: “When it comes to scale and security, mainframes are a tough act to follow. Overall, they’ve been a sturdy and reliable technology for years, despite transactions amounting to millions per day.”
The largest most consistent problem statement raised by the group revolved around the monolithic nature of the workloads that are running on this venerable platform. “We continue to run our most sensitive mission-critical applications on our mainframe,” stated one executive. Others concurred and described how they are looking for ways to leverage investments in technologies like Docker to determine how key applications can be refactored into applications that run in containers on the mainframe. “This would allow us to leverage the robust experience that we've had with mainframes for the past 30, 40 years,” stated another participant.
It is an idea that was shared by several, who noted how the containerization of many mainframe applications that are slated to stay on the platform has reignited interest in the COBOL programming language. As one executive observed: “People -- young people -- are learning COBOL today because they're realizing that there will be continued demand to support those applications for years to come.”
Others noted how the evolving economics of mainframe are making decisions to stay with COBOL-based apps a strong option. “I don't see the mainframe technology or COBOL as a stumbling block because it's a very robust and mature platform. A lot of financial institutions have very expensive applications that were developed in COBOL. These COBOL applications are quite robust. It would cost $300-, $400, $500 million for some of these institutions to refactor.”
Many questions were raised around the direction mainframe pricing is taking as organizations modernize their applications and integrate them into today’s hybrid enterprise infrastructure. “When it comes to mainframe pricing, there needs to be a modernization on the part of the vendor community. The mainframe has to come more in line with pay-by-the-drink models that have been introduced by cloud vendors,” observed one participant. As that point was being made, one of the participants wrote in the chat that there is room in the industry for value propositions that revolve around a “Mainframe-as-a-Service” concept.
Both Aparna and Barry from IBM noted that a significant amount of progress has been made in revisiting how relationships between mainframe technology providers and financial institutions are structured. They noted that solutions are available from IBM to help clients with third-party software companies when the pricing or functionality is not working optimally for clients.
All folks in the room agreed that the ultimate goal is to create a mainframe application modernization strategy that effectively and intuitively integrates all major elements of enterprise infrastructures -- both on-prem and in the cloud. However, while harmonization is critical, refactoring applications is tough. As one participant noted “Refactoring many of these monolithic applications is difficult because they often do about 50 different things. Historically, as the company grew, we acquired more applications that do 50 things. Pretty soon we've got 87 different ways in which we make payments. We have to turn all that into a service, This would ensure that is only one way -- across the institution -- to make a payment.”
We discussed the potential of turning mainframes into intelligent repositories of containerized services. “We have to break those big monolithic applications apart to create services based on a container-type of capability. That's how institutions are going to be able to continue using the robustness of mainframe investments,” noted a participant.
The IBM co-hosts suggested that the best path forward for most financial institutions will likely revolve around the concept of “encapsulation” to develop mainframe modernization initiatives that start with a service interface. “Financial institutions that can encapsulate their applications won’t have to move workloads for them to be containerized.”
NOTE: Encapsulation is a technique that uses object-oriented programming (OOP) to expose the capabilities of an application, without revealing the internal workings of the software program by using an application programming interface (API).
As Barry noted: “Gartner's view of application modernization starts with the simplest, least costly, least risky methods -- which is encapsulation. Encapsulation service-enables applications by using APIs. The other end of the spectrum, from Gartner’s perspective, is to re-write and replace everything, which comes with the most cost and risk. The number one use case we're seeing for containerization on the platform is to containerize and co-locate new workloads that have a high degree of interdependency with the main workload. By running those containerized workloads on the same platform financial institutions can get a tremendous amount of latency and cost benefits.”
Aparna, from IBM, noted that application modernization initiatives have accelerated in the last 18 months, and have typically fallen into scenarios:
The first one revolves around developing mainframe applications that allow financial institutions to leverage new cloud-native features and functionality.
The second approach revolves around performing stringent feasibility analyses on available technology, financial, talent -- and other critical resources -- to determine how workloads can be modernized.
The third category is about hybrid modernization. “I see this category to be the most common scenario. This is when customers in the financial services sector take a hard look at developing an integrated architecture across the mainframe and cloud so that it's more of an interconnected ecosystem,” stated Aparna.
For more information or to inquire about roundtable participation, please visit www.CIO.com.