PropelZ™ enables bidirectional mainframe file system integration: read data directly from VSAM and sequential files, transform it through any processing pipeline, and then deliver it to any target. The input side of the story begins where your data already lives — in native z/OS datasets.
Mainframe data doesn’t start in the cloud. It starts in VSAM files, sequential datasets, tape archives, and GDGs — the file structures that have anchored enterprise computing for decades. Before any analytics platform can process that data, before any modern application can consume it, and before any insight can be drawn from it, the data has to be read from where it actually resides.
This is where PropelZ begins. The VSAM and Sequential File Input connectors are the front door to the entire integration pipeline, turning native mainframe file structures into data that any downstream system can use.
Where the Pipeline Starts
Every transformation, every migration, and every integration scenario depends first on reading the source correctly. Mainframe file systems are not a single, uniform thing — they span keyed VSAM structures, fixed and variable sequential records, generation data groups that roll forward over time, and tape volumes that may hold archives going back years.
Our PropelZ input connectors read these structures natively. Rather than requiring data to be exported, staged, or pre-processed into some intermediate format, PropelZ connects directly to the source files and streams their contents into the transformation pipeline.
This direct-from-source approach matters because it removes the brittle, error-prone steps that traditional integration approaches insert between the mainframe and everything else. There is no separate extract job to schedule, no intermediate file to manage, and no staging layer to maintain. The data moves from its native structure into the pipeline in a single, governed flow.
Reading VSAM Without the Complexity
VSAM files are not simple. They carry key structures, indexing strategies, defined record formats, and access characteristics that vary widely from file to file. Reading them correctly means understanding how each file is organized — key lengths, record layouts, and the relationships between components.
Rather than forcing customers to translate all of that complexity by hand, PropelZ reads VSAM structures using the file’s own definitions. Our connector understands keyed and non-keyed access, handles the record structures it encounters, and presents the data to the pipeline in a consistent form regardless of how the underlying file was built.
For sequential files and GDGs, the same principle applies. PropelZ reads fixed-length and variable-length records, follows generation data group conventions, and coordinates tape volume access using familiar mainframe allocation patterns. Customers describe the source using the same parameters they already know, and PropelZ handles the mechanics of reading it.
Automatic Metadata Discovery
Understanding the shape of mainframe data has traditionally meant maintaining COBOL copybooks — the record layout definitions that describe field names, types, and lengths. Keeping these definitions accurate and synchronized with the actual data is a persistent source of effort and error.
The PropelZ input connectors reduce this burden through automatic metadata discovery. Where the source provides structural information, PropelZ uses it directly, deriving the field-level detail needed to interpret records without requiring customers to hand-build and hand-maintain every definition.
The result is a simpler, more reliable path from raw mainframe records to structured, usable data. Field names, types, and lengths flow into the pipeline as part of reading the source, rather than as a separate maintenance task that has to keep pace with every change.
Beyond Reading: Format Intelligence
Reading data from mainframe files involves more than locating records. Mainframe data uses encodings, numeric formats, and record structures that differ fundamentally from how distributed systems represent the same information. Character data, packed and zoned numeric fields, fixed-length records with specific padding — all of it has to be interpreted correctly before downstream systems can use it.
The PropelZ input connectors include intelligent format handling that bridges these differences automatically. As records are read, the connectors interpret encoding, normalize record structures, and translate numeric and character formats into representations that distributed targets expect.
The goal is making cross-platform integration seamless without requiring customers to understand the low-level differences between mainframe and distributed data representations. Data that starts as a fixed-length record in a native mainframe encoding arrives at its destination in a form the target system can consume directly.
Feeding Any Target
Once data is read from VSAM or sequential files, it enters the same transformation pipeline that powers every PropelZ integration. From there, it can flow to any target PropelZ supports — databases, cloud storage, analytics platforms, and modern applications.
This is what makes the PropelZ input connectors more than a simple file reader. A VSAM file can become structured rows in a cloud database. A sequential dataset can become objects in cloud storage. A tape archive can feed an analytics platform. The source format and the destination format are decoupled, connected only by the transformation pipeline in between.
The same integration patterns that make PropelZ effective for outbound data movement now work from the source inward, enabling truly bidirectional mainframe integration.
Closing Integration Loops
These input capabilities reflect real customer requirements. Organizations modernizing their data architectures need to read existing mainframe files reliably as the starting point for migration, analytics, and application integration. Others need to incorporate mainframe file data into hybrid workflows where modern systems process information that originated on z/OS.
Reading from VSAM and sequential files is the natural complement to writing back to them. Together, the PropelZ input and output connectors let data make the complete round trip — read from native mainframe structures, transformed through any pipeline, and returned to mainframe formats when needed. Instead of building custom solutions or accepting one-way data movement, customers can use PropelZ for the entire journey in both directions.
Architectural Integration
The input connectors integrate seamlessly with PropelZ’s transformation pipeline. Any target that PropelZ can write to — databases, cloud storage, APIs, other mainframe files — can now be fed directly from native mainframe file sources.
This means complex transformation workflows can begin with mainframe file reads. Start with a VSAM file or sequential dataset, apply business rules through PropelZ transformations, then write the results to a cloud database, an object store, or a modern application target.
The input and output connectors are mirror images of one another, working from opposite ends of the same pipeline to give PropelZ a complete, symmetric integration model.
Enterprise Integration Maturity
The pairing of comprehensive input and output capabilities represents a mature approach to enterprise integration. Reading from the mainframe and writing back to it are two halves of the same discipline, and treating them as a unified capability rather than separate tools reflects how sophisticated integration architectures actually work.
As organizations move beyond one-way migration toward ongoing, bidirectional integration, the ability to read native mainframe files reliably becomes as important as the ability to write them. Some data needs to flow off the mainframe. Some needs to flow back. Some needs to make the complete round trip for validation or hybrid processing. All of it begins with reading the source correctly.
Beyond Migration: Integration Strategy
With comprehensive input and output capabilities, PropelZ functions as a true integration platform rather than a migration tool. Organizations can design data architectures that leverage the strengths of both mainframe and distributed systems without being constrained by platform boundaries.
Mainframe systems hold data of unmatched reliability and depth. Modern analytics platforms excel at processing that data at scale and applying advanced techniques to it. Instead of choosing between these strengths, enterprises can read mainframe data directly into modern pipelines, process it wherever it provides the most value, and return results wherever they are needed.
The result is integration architecture that adapts to business requirements rather than forcing business processes to adapt to platform limitations. Data flows where it provides the most value, processed by systems optimized for specific tasks, then shared seamlessly across the enterprise.
This represents the true promise of hybrid computing: leveraging the best capabilities of every platform while maintaining the integration flexibility to evolve as requirements change.
Next Steps
- Schedule a PropelZ briefing.
- Watch a demo of PropelZ.
- Try PropelZ.
Learn More
- Visit our Customer Briefing Center.
- Review all VirtualZ Use Cases, Thought Leadership papers, and Solution Briefs.
- Explore our YouTube channel, podcasts, and blog.
- Still have questions? Contact us.
