DevOps is the standard for delivering speed, quality and continuous improvement as enterprises move toward digital transformation. Although cloud-native and distributed platforms have widely adopted DevOps practices, mainframe applications often remain isolated, operating under legacy delivery models that limit enterprise agility.
This blog outlines how DevOps was successfully extended to a large portfolio of mainframe applications, aligning technical modernization with clear business outcomes in one of my projects.
The Challenge: Legacy Delivery Meets Modern Expectations
The organization depended on mainframe applications for core business processing, yet these systems were delivered using:
- Waterfall-style release cycles with long lead times
- Manual build and deployment processes
- No Git-based source control for mainframe code
- Limited automation and visibility into changes
- No security or vulnerability scanning for mainframe applications
In contrast, their modern platforms had mature CI/CD pipelines. This disparity created operational silos, slowed innovation and increased delivery and security risks.
The Approach: Modernizing Delivery Without Replacing the Mainframe
The objective was not to replace the mainframe, but to modernize how it participated in the software delivery life cycle.
We devised a solution that focused on integrating mainframe environments into existing enterprise DevOps tooling while preserving platform stability.
Technical Architecture: Enabling DevOps for the Mainframe
Bridging DevOps Tools With the Mainframe
Zowe CLI was introduced to enable modern CI tools, such as Jenkins, to interact directly with the mainframe machines (LPARS). Jenkins pipelines invoked Zowe CLI commands to execute build and compile operations.
These commands were routed through Endevor web services configured within the mainframe environment. This required a one-time foundational setup involving:
- Setting up the Jenkins master-slave configuration
- Installing the Zowe client CLI
- Enabling Endevor services in mainframe applications
- Enabling security and user configurations to access the mainframe from Zowe
Once in place, the integration enabled automated, programmatic trigger of mainframe processes for CI/CD.
Source Control and Continuous Integration
Endevor Bridge for Git enables bi-directional synchronization between the mainframe source code and Git repositories.
This allowed:
- Version-controlled code in Git
- Standard Git workflows
- Automated CI triggers
Interestingly, it allowed developers who preferred the legacy style of working on mainframe green screens to continue working in a non-disruptive fashion.
File additions or modifications in Git initiated Jenkins pipelines, which compiled and validated the corresponding artifacts on the mainframe, completing the full CI cycle.
Security and Compliance Integration
With mainframe code available in Git, enterprise-grade tooling could now be applied uniformly:
- Static code analysis
- Vulnerability and security scanning
- Compliance and governance checks
This eliminated long-standing security blind spots and brought mainframe applications in line with organizational risk and compliance standards.
Business Outcomes: From Intent to Impacts
The integration of DevOps into the mainframe environment delivered measurable business value:
- Reduced release cycles, enabling faster time to market
- Improved delivery predictability through automation
- Enhanced security posture with standardized scanning and controls
- Unified delivery model across legacy and modern platforms
- Strong alignment between mainframe and digital engineering teams
The mainframe application delivery processes transitioned from being a perceived bottleneck to becoming an active participant in CI/CD.
Key Takeaway
DevOps processes should not be confined to modern stacks alone. The right architectural approach can help adopt and enable modern delivery practices on legacy platforms as well.
Mainframe systems have always been known for their reliability and resiliency. During this initiative, we integrated mainframe systems into enterprise DevOps pipelines and enabled greater speed and end-to-end agility in their delivery cycle as well.