mvsMF

mvsMF is a implementation of the z/OSMF REST API for the classic MVS 3.8j. The project provides essential functionality of z/OSMF, focusing on key endpoints for dataset and job handling. This allows users to use modern clients such as the Zowe Explorer for Visual Studio Code (or JetBrains IDEs) and Zowe CLI to interact with their host datasets, submit jobs, and view job outputs on MVS 3.8j.

This project is implemented as a CGI module for Mike Rayborn's HTTPD server and relies on his CRENT370 libraries for efficient system interaction. A huge thanks goes to Mike Rayborn for his invaluable contributions to the MVS community.

Features

• Dataset Handling: Access, create, update, and delete datasets.
• Job Management: Submit jobs, check their status, and retrieve their output.

Installation

Note: This project is currently a Work in Progress and is not intended for production use.

Prerequisites

• MVS 3.8j system
• Mike Rayborn's HTTPD version >= 3.3.0 installed and configured

Steps

1. Transfer the provided XMIT file to your MVS system.

2. Restore the XMIT file to create the necessary load module:

   RECEIVE INDATASET('your.xmit.dataset') 

3. Copy the the load module from the received dataset into the LINKLIB dataset used by your HTTPD server.

4. Update your HTTPD configuration file to include the following entry:

   cgi.MVSMF="/zosmf/*"

5. Restart the HTTPD server.

Building mvsMF

mvsMF uses mbt (MVS Build Tools) for cross-compilation, assembly, and linking on MVS.

Prerequisites

• An MVS 3.8j system reachable via IP (TK4-, TK5, MVSCE, or similar)
• The c2asm370 cross-compiler
• Python 3.12+

Quick Start

git clone --recursive https://github.com/mvslovers/mvsmf.git
cd mvsmf
cp .env.example .env    # edit with your MVS connection details
make bootstrap          # resolve dependencies, allocate datasets
make build              # cross-compile and assemble on MVS
make link               # link-edit into load module

Make Targets

Target	Description
make bootstrap	Resolve dependencies, allocate MVS datasets
make build	Cross-compile C sources and assemble on MVS
make link	Link-edit all modules into a load module on MVS
make install	Copy load module into the target LINKLIB on MVS (not yet implemented)
make package	TRANSMIT load library to XMIT format and download
make compiledb	Generate compile_commands.json for clangd
make clean	Remove generated .s and .o files
make distclean	Full clean including mbt cache and stamps

Configuration

Project structure is defined in project.toml. Local MVS connection settings go in .env (never committed). Copy .env.example to get started.

Key variables:

Variable	Description
MBT_MVS_HOST	IP or hostname of the MVS system
MBT_MVS_PORT	mvsMF API port
MBT_MVS_USER	MVS userid
MBT_MVS_PASS	MVS password
MBT_MVS_DEPS_VOLUME	Volume for RECEIVE (see note below)

See .env.example for the full list.

MVS/CE users: MVS/CE systems typically have no public volumes available for TSO RECEIVE. Set MBT_MVS_DEPS_VOLUME in your .env to a work volume (e.g. PUB000) so that dependency datasets are placed on a known volume.

Usage

Once installed, you can use tools like Zowe CLI or IDE plugins like Zowe Explorer to connect to your MVS 3.8j instance. You'll be able to:

• List datasets
• Submit and monitor jobs
• Retrieve and view job outputs

Acknowledgments

This project wouldn't have been possible without the incredible work and support of Mike Rayborn. His HTTPD server and CRENT370 libraries are at the core of this implementation. A huge thank you for your contributions to the MVS community!

Contributing

Contributions are welcome! If you're interested in helping with development, testing, or documentation, feel free to open an issue or submit a pull request.

License

This project is licensed under the MIT License

Disclaimer: This project is still under active development and is not ready for production use. Use it at your own risk and report any issues or feedback to help improve it.