In this series, we shine a light on the people who make the Ansible® for IBM® Z® community thrive. We share the inspiring stories of users who combine experience, skill, and curiosity to shape strategies, influence organizations, and take our tools to new heights.
Prasanna Venkatesh Balasubramanian is a Lead Software Engineer at M&T Bank with over 20 years of mainframe experience across India and the US. He is passionate about modernizing mainframe systems and is currently driving DevOps transformation on IBM Z through Ansible-powered CI/CD automation.
Michael Wright is an Engineering Manager at M&T Bank with over 30 years of Mainframe & Distributed software development, managing teams in Ireland, Scotland, Italy, Canada, and the US, supporting very large systems while also working internationally. While at M&T, he was instrumental in educating the company’s core banking Mainframe development teams on the fundamentals of the Agile framework and is currently overseeing the two-year journey of deploying M&T’s transformational hybrid DevOps on Z solution, enabling the automation capabilities & process improvements the platform will deliver.
How did you get involved in the work that you do now?
Prasanna: I joined M&T Bank five years ago. Modernizing our systems is a significant initiative, and I became actively involved in the DevOps on Z project, which Mike started. This project has been a truly exciting journey, and we recently went live. Our developers are now actively using the platform to deploy code to production, which represents a significant milestone for our team.
How did you initially get involved in the DevOps on Z project?
Prasanna: When I first joined the team, Mike introduced me to a related project. I quickly became engaged—writing Java code, managing Ignite, and deploying onto Z systems. It was a fantastic learning opportunity. When Mike started the DevOps on Z project, he enlisted my help, and it has been a collaborative and rewarding journey ever since.
What is your role in the DevOps on Z project now, and have you faced any challenges?
Prasanna: I'm focusing on designing, architecting, and providing solutions for the platform. A significant challenge we encountered was that many DevOps tools we wished to use were not initially compatible with the mainframe environment. What works seamlessly on distributed systems does not always translate easily to mainframe systems, and this discrepancy required us to experiment, learn, and adapt. The learning curve was steep, but it was an enriching experience.
What tools and technologies did you implement in this project?
Prasanna: We employed a range of tools to create our CI/CD pipeline. GitLab serves as our primary source control and continuous integration (CI) tool. We integrated Ansible to connect GitLab with our mainframe deployments, which allows us to SSH into the mainframe to deploy the executables coming out of DBB (Dependency-Based Build) into our target libraries . We incorporated SonarQube and established custom user interfaces using tools like IDz to manage and visualize our pipeline effectively.
Mike, what is your current role, and what led you to initiate the DevOps on Z project?
Mike: I'm currently Hogan Manager at M&T Bank, where I oversee our core banking system that supports thousands of interfaces. I’ve been with M&T for about six years, managing a team that includes both developers like Prasanna and business analysts.
Shortly after I came on board, I was tasked with assessing various DevOps solutions for the mainframe. DevOps was still emerging in the mainframe space, so it was a new frontier for many of us. Based on my experience with distributed systems and our requirements on the mainframe, I designed the high-level architecture for implementation. Once we received executive approval, I transitioned into the project manager role to oversee the implementation.
What does the solution look like today?
Mike: We have one of the most comprehensive DevOps on Z implementations in the banking industry. Our integrated toolset includes Git and GitLab, DBB, SonarQube, Jira and ServiceNow for issue and change management, and Ansible for deployment automation.
A notable achievement was transitioning from our 30-year-old source code management system, Panvalet, to Git. This was a massive leap forward, and not many organizations have undertaken such a transformative step in one go.
Why was automation such a key focus for this project?
Mike: Historically, mainframe automation capabilities lag those in distributed systems. Our goal was to change that dynamic significantly. We use GitLab for source control and continuous integration, DBB to manage our mainframe builds, and SonarQube to analyze and maintain code quality. Ansible drives our continuous delivery (CD) processes.
Initially, we considered UrbanCode Deploy, but after consulting with our distributed teams, we pivoted to Ansible. We found that Ansible offered greater flexibility, was more future-proof, and fit perfectly with our automation vision.
How did you implement Ansible and prepare your team for this transition?
Mike: We partnered with experts from Red Hat and IBM, including Jacob Emery, who conducted a three-day workshop for our developers. This workshop proved invaluable in helping us kick off our transition.
Prasanna has been a critical driver of the initiative. His deep knowledge of COBOL and Python skills has made him invaluable, especially since Ansible is based on Python. His passion for automation has pushed the initiative forward— he has actively worked with Red Hat and IBM to help organize workshops, and he has mentored team members through the transition.
How have you prepared the rest of your team for this shift?
Mike: I encourage our COBOL developers to learn Python, which will enable them to contribute to automation efforts using Ansible. It’s a significant cultural and technical shift, but it’s absolutely necessary for the future of mainframe development.
Where does the project stand currently, and what’s next for M&T Bank?
Mike: We recently went live with the new platform. We’ve successfully transitioned off our legacy source code management system, and we’re using Git and GitLab for both CI and CD processes. Ansible is managing the deployment end-to-end.
Hogan is the first mainframe application at M&T to launch on this new platform. Ultimately, our goal is to migrate all approximately 100 mainframe applications at M&T onto this new platform. We’re at the cutting edge of mainframe DevOps, and while this journey presents challenges, it’s incredibly exciting. Our implementation has been recognized by IBM and others as one of the most advanced in the industry.
How has this hybrid approach changed the way your team operates?
Prasanna: We’ve always had some level of automation within the mainframe through batch jobs, scripts, and specific tools. However, the platform we’re developing allows us to automate in much more dynamic and user-friendly ways, enabling us to build interfaces, create surveys, and streamline workflows that weren’t feasible before.
Could you share an example of how this new strategy is being implemented?
Mike: Absolutely! One challenge we faced was getting our business analysts familiar with tools like Git. Since they’re not developers, we needed to simplify the process for them. Prasanna identified the need and proactively addressed it—he took the initiative to create a straightforward web interface using Python that abstracts away the complexity, enabling analysts to perform tasks like merges and updates without needing to learn Git commands. This forward-thinking solution illustrates the kind of contributions he brings to the team and is shaping how the platform evolves and scales across the organization.
Was there any resistance to changing direction at any point?
Mike: Not really. We enjoyed a lot of trust from senior management. They empowered us to make technical decisions as long as we acted thoughtfully and thoroughly. I organized a cross-functional group involving our consultants, Prasanna, the application team, and some of the distributed folks to deliberate on the merits of switching to Ansible.
What ultimately convinced you to adopt Ansible?
Mike: First, Ansible was already in use at M&T, so we didn’t have to invest in a new license. More importantly, it aligned perfectly with our primary goal: increasing automation. Ansible provided us the flexibility and scalability we needed for effective deployment automation and other processes.
How do you feel about that decision?
Mike: It was absolutely the right move. Ansible has become a cornerstone of our DevOps platform, enabling us to automate not only deployments but a vast array of tasks across our operations. It has allowed us to develop a more modern, scalable, and efficient solution that serves both our mainframe and distributed environments.
Can you share a specific use case where Ansible is aiding your pipeline management and CI/CD processes?
Prasanna: We’re exploring integrating Ansible with ServiceNow for change-driven deployments. Our goal is to have ServiceNow automatically trigger deployments. When a change request is approved and the implementation window opens, ServiceNow will send a webhook to Ansible, which will execute the relevant playbook to deploy the changes.
That sounds powerful! How are you managing deployments in the interim?
Prasanna: We’re using CA-7 as a temporary solution. Our GitLab pipeline hands the deployment list to Ansible, which prepares the necessary staging libraries. From there, we initiate a deployment job using a template and employ the CA-7 Ansible collection to schedule it. We connect to ServiceNow to retrieve the change number and implementation timeframe to set a hold on the CA-7 job, ensuring that no deployments occur before the approved window.
So even in this interim state, Ansible is playing a critical role?
Prasanna: Exactly! Without Ansible, we would not have bridged that gap so effectively. Ansible has enabled us to create robust guardrails around our deployment processes. Once the ServiceNow integration is complete, we’ll transition to a fully automated model, where Ansible will manage everything—from copying deployments to target libraries to executing new CICS copies using the Ansible CICS collection.
We heard that your team is preparing for a sysplex split. How do you envision Ansible playing a role in that transition?
Prasanna: We’re anticipating a sysplex split between our development and production systems. While we haven't yet tested this, I firmly believe that Ansible will be crucial. It should facilitate connections to multiple hosts, allowing seamless component transfers—something that would otherwise be complex and error-prone if managed through multiple jobs manually.
Prasanna, it’s clear that you’re leveraging Ansible in various impactful ways. Could you have accomplished all of this without Mike’s support?
Prasanna: Not at all! Mike has been pivotal in navigating various challenges. Whenever we’ve approached him with our needs, he’s been the one to clear the path and make things happen. His support has been instrumental in ensuring that we can focus on building the solution.
Mike: I aim to do my part by supporting the team, removing obstacles, and ensuring that they have everything they need. Prasanna is the architect here; my role is to enable him to build.
Mike and Prasanna, what would be your message to organizations that perceive the mainframe as legacy technology?
Mike: I advise organizations to take a fresh look at what’s happening in the mainframe space. For years, the misconception has been that mainframes are outdated, but modern mainframe technology supports various languages and frameworks and is increasingly capable of handling tasks typically associated with distributed systems. The reality is that we’re doing things on the mainframe today—like streaming data, hybrid DevOps, and event-driven automation—that were once thought to be exclusive to distributed systems.
I would also tell them to prioritize automation. If automation is a priority for your DevOps strategy, Ansible is an essential tool. Its powerful capabilities can genuinely revolutionize how you manage deployments and operations.
Prasanna: Tools like Ansible are integral to bridging this gap. Without Ansible, we wouldn’t have achieved the level of automation we have today. It’s empowering us to modernize our workflows, integrate with platforms like ServiceNow, and prepare for a future characterized by event-driven and intelligent automation. Ansible has truly become our core deployment engine. It has enabled us to standardize and automate processes that were previously manual or disjointed. Despite some limitations, it is a powerful tool helping us evolve at scale.
How has the partnership with IBM and Red Hat been throughout this journey?
Mike: The support from IBM and Red Hat has been outstanding. They’ve gone above and beyond—extending extra workshop sessions, assisting us in running sessions on our own Ansible Automation Platform, and ensuring our developers felt supported throughout the transition. I can confidently say that our partnership has been fantastic. IBM and Red Hat genuinely want us to succeed, and we are incredibly appreciative of their ongoing support.
Stay up to date with the Ansible User Spotlight Series and Ansible for IBM Z community
Join the IBM Z & LinuxOne community topic group for Ansible for IBM Z. If you want to join our Ansible for IBM Z Community Guild monthly calls, where the IBM Z team and Ansible customers share roadmaps on content, use-cases, and collaborate on Ansible adoption, fill out this interest form.
Check out the last User Spotlight Series article here!
Here are some additional resources if you are interested in getting started automating with Ansible for z/OS:
Thank you to the team behind the Ansible for IBM Z User Spotlight Series!