Mike Thompson
In the realm of modern software development, the principles of DevOps have become synonymous with agility, efficiency, and collaboration. Yet, when it comes to mainframe environments, embracing DevOps practices presents a unique set of challenges that demand innovative solutions.
DevOps on mainframe?
Mainframes, with their robust architecture and heritage application and infrastructure estate, present a challenge on the path to implementing seamless DevOps pipelines across an organisation. However, overcoming these hurdles can unlock substantial benefits and productivity gains via the integration of non-mainframe tooling.
The mainframe conundrum
Mainframes have long been the backbone of critical business operations, supporting vast amounts of data and processing power. However, their traditional development methodologies often rely on rigid, sometimes monolithic structures that may hinder rapid iteration and deployment. The transition to a DevOps model requires a fundamental shift in mindset and processes, which can be particularly daunting in the mainframe environment and for traditional development teams.
One of the primary challenges stems from the complexity of mainframe applications and the intricate interdependencies among various components. Unlike modern cloud-native architectures, mainframe systems often lack the flexibility and modularity required for seamless integration into automated DevOps pipelines.
Navigating heritage landscapes with DevOps
Introducing DevOps onto the mainframe platform requires a strategic approach that balances the need for innovation with the constraints of heritage systems. One key strategy involves leveraging emulation and abstraction layers to bridge the gap between mainframe environments and modern DevOps toolchains. By encapsulating heritage functionalities within containerised environments, organisations can integrate mainframe applications into CI/CD workflows seamlessly.
Furthermore, the adoption of infrastructure-as-code (IaC) principles enables teams to manage mainframe configurations programmatically, streamlining provisioning and deployment processes. Tools like Ansible offer declarative approaches to infrastructure management, abstracting the complexities of mainframe administration and promoting consistency across environments.
Benefits of non-mainframe tooling in DevOps
While mainframe environments present unique challenges, the incorporation of non-mainframe tooling offers significant benefits and productivity gains through automation. By embracing cloud-native technologies and open-source frameworks, organisations can unlock new levels of agility and efficiency in their development workflows. These tools can integrate into traditional Software Configuration & Application Lifecycle Management tooling, such as the ever-popular Endevor from Broadcom, and existing processes.
Continuous integration and deployment
Non-mainframe CI/CD tools can provide robust automation capabilities that streamline the build, test, and deployment processes. By orchestrating workflows across heterogeneous environments, these platforms enable rapid feedback cycles and accelerate time-to-market for mainframe applications. Orchestration tools, such as Jenkins, allow the scripting of processes to fully automate the mainframe DevOps pipeline and give ‘power to the elbow' for Developers and Infrastructure teams alike. Automated regression/unit testing, using tools such as Topaz for Total Test (now BMC AMI DevX Total Test), is a must-have in any mainframe Application-based DevOps pipeline and may provide the biggest ‘bang-for-your-buck'.
Continuous monitoring and observability
Effective monitoring and observability are essential components of any DevOps initiative, providing real-time insights into system performance and reliability. With the proliferation of open-source monitoring solutions such as Grafana and AI-Ops, organisations can gain visibility into mainframe workloads and proactively identify bottlenecks or anomalies.
Conclusion
In conclusion, while introducing DevOps on mainframe platforms poses formidable challenges, the benefits and productivity gains offered by automation and non-mainframe tooling far outweigh the complexities. By embracing innovative strategies and leveraging cloud-native technologies, organisations can navigate the heritage landscape and unlock new levels of agility and efficiency in their mainframe development workflows. As the digital landscape continues to evolve, embracing DevOps on mainframe will be essential for organisations seeking to remain competitive in an ever-changing marketplace.
I'm Mike Thompson, global head of mainframe developer engineering at HSBC. I've spent a career working on and around the mainframe platform, mainly as a COBOL, CICS and DB2 Developer of over 25 years. I gained a First Class Honours degree in Computer Science at Teesside Polytechnic, following five years as a Mechanical Engineering Draughtsman at British Steel. My first IT role was at a County Council, then moved into the private sector at Midland Bank, later HSBC, Experian and HBoS, and as a consultant at HSBC, Santander, Caterpillar and Royal Mail. I currently run a global team across seven countries.
