Mainframe Meets DevOps: Git Integration
Key Points
- DevOps (often called DevSecOps, Biz DevSecOps, QA Ops, etc.) is about unifying all development, security, and operations teams, and this unification must include mainframe environments.
- Traditional mainframe deployment uses library managers/production‑control tools that handle source promotion much like modern CI/CD pipelines manage code and artifact flows in the distributed cloud world.
- Migrating mainframe source (COBOL, PL/1, JCL, REXX, DDL, CICS definitions, etc.) into Git—using Git attributes to specify the correct z/OS code page—places it alongside cloud‑native code and infrastructure‑as‑code assets.
- Once in Git, the mainframe build can be modernized: automated scans, unit tests, security checks, and artifact creation are added, and the resulting binaries are stored in a shared artifact repository.
- This unified, Git‑centric pipeline delivers the same rapid feedback, consistency, and automation benefits to mainframe workloads that DevOps provides for cloud‑native applications.
Sections
- Mainframe Integration with DevOps - The speaker outlines how DevSecOps principles—centralized pipelines, source control, automated quality and security checks, and promotion workflows—are applied to mainframe development using library managers and production‑control tools, paralleling modern Git‑based cloud‑native practices.
- Modernizing Mainframe Code with Git - The speaker describes moving COBOL, PL/1, JCL, REXX, and related assets into Git—using .gitattributes for code‑page conversion—to enable a unified, language‑agnostic build pipeline with automated testing and artifact repository integration.
- Automating Ticket Creation for DevOps - The speaker introduces modern automation techniques to eliminate tedious tasks such as ticket creation, enabling smoother, collaborative DevOps workflows on the mainframe.
Full Transcript
# Mainframe Meets DevOps: Git Integration **Source:** [https://www.youtube.com/watch?v=umo6kKXD0-Q](https://www.youtube.com/watch?v=umo6kKXD0-Q) **Duration:** 00:06:48 ## Summary - DevOps (often called DevSecOps, Biz DevSecOps, QA Ops, etc.) is about unifying all development, security, and operations teams, and this unification must include mainframe environments. - Traditional mainframe deployment uses library managers/production‑control tools that handle source promotion much like modern CI/CD pipelines manage code and artifact flows in the distributed cloud world. - Migrating mainframe source (COBOL, PL/1, JCL, REXX, DDL, CICS definitions, etc.) into Git—using Git attributes to specify the correct z/OS code page—places it alongside cloud‑native code and infrastructure‑as‑code assets. - Once in Git, the mainframe build can be modernized: automated scans, unit tests, security checks, and artifact creation are added, and the resulting binaries are stored in a shared artifact repository. - This unified, Git‑centric pipeline delivers the same rapid feedback, consistency, and automation benefits to mainframe workloads that DevOps provides for cloud‑native applications. ## Sections - [00:00:00](https://www.youtube.com/watch?v=umo6kKXD0-Q&t=0s) **Mainframe Integration with DevOps** - The speaker outlines how DevSecOps principles—centralized pipelines, source control, automated quality and security checks, and promotion workflows—are applied to mainframe development using library managers and production‑control tools, paralleling modern Git‑based cloud‑native practices. - [00:03:02](https://www.youtube.com/watch?v=umo6kKXD0-Q&t=182s) **Modernizing Mainframe Code with Git** - The speaker describes moving COBOL, PL/1, JCL, REXX, and related assets into Git—using .gitattributes for code‑page conversion—to enable a unified, language‑agnostic build pipeline with automated testing and artifact repository integration. - [00:06:13](https://www.youtube.com/watch?v=umo6kKXD0-Q&t=373s) **Automating Ticket Creation for DevOps** - The speaker introduces modern automation techniques to eliminate tedious tasks such as ticket creation, enabling smoother, collaborative DevOps workflows on the mainframe. ## Full Transcript
DevOps and the mainframe.
They go together.
But what is DevOps?
When we think about DevOps, we really think about DevSecOps.
And DevSecOps, Biz DevSecOps, QA Ops.
All of these various names are used for DevOps,
but it really is, "Let's bring everybody together",
and that's why the mainframe and DevOps has to go together.
We think about cloud native development,
we have pipelines for development,
we use Git as our source manager for development
when we're doing cloud native activities.
We think about the mainframe instead, we have library managers,
or, as many times they're called, production control tools.
And they were created to control the process
of deployment into production,
which is what pipelines do in the distributed side as well.
In many ways they manage the source code so your source is managed.
But in order to move between environments,
you use this simple promotion process.
All you're doing is managing source and your loads through this process
and someone else is doing this promotion to production.
So you have to work with someone else to get through this process.
And it manages your COBOL and your PL/1
And if you're lucky, you're REXX and your JCL.
But in the distributed world in my Git pipeline,
I have Git as my source code manager.
And then the pipeline does things such as scanning my source.
So I get code checks, I get code quality checks,
I can get security scanning, I can do my build,
I can run my unit tests, automated unit tests, not over here.
I'm getting my automated checking right away.
When I'm doing my development, I get my deploy
and I can run more automated test.
So here I'm getting all this value out of the pipeline.
with storing my source in Git. And normally
it's Java and other languages,
but it's also things like TerraForm Patterns.
So not only is my source code
for my application here, but my infrastructure definition.
Yeah, that's missing over in this side.
I don't have that in this world.
So if we think about the mainframe, DevOps, these pipelines, this transformation,
why is my source over here?
Why am I doing things different?
I don't have to ... Git,
it can manage COBOL or PL/1 just as well.
So what do I need to do?
Get my source
and move it into Git.
That's the first step.
Take that source code out
and there's a nice capability of the Git attributes file
In the Git attributes file,
I can indicate what code page I need in the z/OS environment
and it can be UTF8 sitting in Git.
And so my source code can be there alongside everything else.
And now my COBOL and my PL/1
and everything else, JCL, REXX, anything else,
but also my DDL or my CICS definitions.
They can all sit in this Git environment to be then deployed.
I pull the build process out of the system
and convert it into a modern build process.
I can add in my unit tests that I didn't have before
and the deploy, I pull that out
and I can deploy and put my artifacts in an artifact repository
just like they are over here.
And so I have this consistent way of working,
whether or not I'm doing COBOL PL/1 or Java.
It's all the same as a developer.
And what's even better when we think about multi language
capability, if I'm doing CICS, there's no reason
I can't have Java and COBOL
or other languages working together in my CICS environment
and now they can be together in my source code manager.
As a developer, I get to do work one way in one source
code manager using one pipeline and get all the advantages
of the automated testing capabilities that are there.
All of the checking code, scanning code quality capabilities,
all of it provided to me by simply moving into this environment.
Now, I said simply,
I know this isn't simple change.
I know this is a transformation.
I know this is hard to move
from the way you've been doing work for a long time.
But think about it.
This is the way everything else is done today.
We are developers.
We need to be part of this continuous learning culture
and take on this new way of working and so that the mainframe
is just as modern and can work
in all of the same ways as everything else.
And once I've done this, now I can take advantage
of all those other capabilities available on the system,
such as, you know, bringing in a little AI into the system.
So hopefully this has given you an introduction to it's time to move.
Let's take advantage of the modern ways of working the capabilities
we can do through automation to remove those extra tasks that waste
and all those things that we hate doing, like creating tickets
and get to work together across the organization.
Thanks for
watching and don't forget to click like and subscribe.
And to find out more about DevOps and the mainframe, check out the links below.