Solutions / Mainframe DevOps

Get CA Librarian and Panvalet source into Git — and finally into a real build

Librarian and Panvalet have safely stored your source for decades, but they were never build systems and never will be. We extract your members and their history, model them in Git, and add the impact-aware build these libraries never had.

The Problem

A library is not a pipeline

CA Librarian and CA Panvalet do one thing well: they keep versioned source members, with archived deltas and a change history, inside a proprietary master file. That was modern in the 1970s and dependable ever since. But they are storage, not delivery. There is no branching, no dependency-driven build, and no promotion model — the member you check out is the member you compile, and working out what else that change affects is a manual exercise.

That leaves Librarian and Panvalet shops a step further back than Endevor shops. Where an Endevor migration is about preserving a promotion path, a Librarian or Panvalet migration is about gaining capabilities you never had: real version control, and a build that knows a changed copybook forces its dependents to recompile. IBM Dependency Based Build (DBB) provides exactly that — it scans your source to construct the dependency graph, recompiles what a change actually touches, and re-links statically bound load modules.

These are also among the oldest products in Broadcom’s CA mainframe portfolio, carried forward with minimal investment. The developers who knew them are retiring, and their replacements expect Git — not a ++INCLUDE and an ISPF library panel.

How The Migration Runs

The migration, in five stages

Because there’s no promotion lifecycle to preserve, the hard part isn’t untangling process — it’s cleanly extracting decades of members and reconstructing the dependency knowledge the library never tracked.

01
Extract members and history from the master file
We pull every member out of the Librarian master file or Panvalet library — source, JCL, copybooks, and the archived version history — in a clean, verifiable batch, without hand-editing a single member.
02
Model it in Git
Members land in a Git repository organized the way your teams actually work — with real branches and pull-request review — instead of a flat library keyed by member name. Where useful, the archived version history is preserved as commit history.
03
Reconstruct dependencies with DBB
Librarian and Panvalet never recorded what depends on what. DBB builds that graph from the source itself — resolving includes and program-to-copybook relationships — so an impact build recompiles the dependents of a change and re-links the load modules that statically bind them.
04
Wire up deployment
Build output needs to move. We package promotion using IBM Wazi Deploy for scripted, Git-native z/OS deployment, or IBM DevOps Deploy where one workflow-driven tool has to span mainframe and distributed platforms — so deployment becomes an auditable step, not a manual copy.
05
Train your developers on the workflow
Your team learns branching and pull-request review against the code they already maintain — so the move off Librarian or Panvalet is also the moment they stop working in isolation from the rest of engineering.
What You Gain

Capabilities the library never offered

Impact-aware builds

A changed copybook rebuilds its dependents automatically — the analysis Librarian and Panvalet left entirely to the developer.

Real branching & review

Feature branches and pull requests replace member-level checkout, bringing mainframe changes into the same review discipline as the rest of your code.

One toolchain

Your z/OS source lives in the same Git, CI, and deployment tooling as everything else — no separate island only a shrinking group knows how to run.

Why Strongback

Extracting clean source from a Librarian master file or a Panvalet library is its own craft. We’ve done it — and built the DBB pipeline on the other side, not just dumped members into a repo.

— Strongback Consulting, mainframe DevOps since before it had a name
Next Step

Tell us what’s in your libraries. We’ll tell you what moves first.

A migration assessment inventories your Librarian or Panvalet members and maps them against a Git + DBB model — no commitment, no generic sales deck.