The Linux Foundation Projects
Skip to main content
Blog | GenevaERS

GenevaERS updates

By | March 11, 2025

Written by Andrea Orth, Lead for the GenevaERS 

GenevaERS rapidly produces many outputs while efficiently reading large volumes of data only once. It is extensible, and adding more reports does not affect the performance. These are easily developed using a graphical interface. GenevaERS rapidly produces many outputs while efficiently reading large volumes of data only once.

Research and development – Inter-language framework

Below is a bulleted list of characteristics of the framework proof of concept (POC). The list begins with the JVM only being started once. Supported applications can be single or multi threaded. The supported mainframe languages are Assember, C++, COBOL, PL/1. Success criteria is proving viability and run-time efficiency.

Areas of focus were:

  • z/OS environment
  • Run time efficiency
  • Multi threading
  • Multi language
  • Starting the Java Virtual Machine (JVM) only once
  • Prove the viability calling Java from Assembler language or a third generation language such as COBOL or C++

After the Java daemon starts the JVM, it starts the Assembler program. The program performs an INIT call telling the Java daemon how many Java threads need to be created in the Java thread pool. A Java thread is created for each subtask under the Assembler program. Each method call by the Assembler program is dynamically loaded and executed in the Java thread call.

The data being passed between the mainframe language and Java can be either 31 bit or 64 bit addresses. 

The findings of the POC were:

  • The overhead of the Assembler program calling Java was surprisingly low
  • Data transmission between application is possible
  • Multiple levels of interfacing is possible
  • Java Batch Launcher and Toolkit (JZOS) was integral for the success of the POC

Technology advancements

Mainframe module MR88 was refactored for easier maintenance and performance was improved by using a hashing algorithm. Functions COUNT and AVERAGE were also added.

In 2024, GenevaERS was installed at a European organization which stores data in ADABAS. GenevaERS was modified to read ADABAS. The ADABAS enhancement will be completed in 2025 with the final changes made in the user interface.

2025 will begin exploration and prioritization of a variety of software enhancements. The project will begin researching moving metadata definition storage away from relational databases to an agnostic solution using a file format such as json, yaml, or xml. The metadata definitions could be stored in a modern source code management system such as GitHub. Definition changes could be versioned and packaged in a release with other related code changes.

What the new mainframe will enable

Anyone interested in GenevaERS will have the ability to run the demo. Additional use cases will be created so a person can walk through the documented steps and run GenevaERS. 

A college student expressed interest in the GenevaERS project. Currently he is making small non-mainframe changes to gain experience with Git / GitHub, but his interest lies with the mainframe. He will have the opportunity to run the GenevaERS batch phase, view standard and GenevaERS reports sent to the system for review and debugging purposes.

Documentation

The documentation site was created using GitHub pages and Jekyll. Changes were made to the presentation of information to aid a new user or someone curious about GenevaERS. The site landing page provides links to installation and setup of the GenevaERS demo along with links to basic concepts and skills.

The step by step instructions walk a new user through making code changes and running the batch phase to create output. New demo user cases created next year will also be documented on the website.

GenevaERS.org website

2025 will bring a new version of the GenevaERS website, moving to a modern solution utilizing GitHub pages instead of WordPress. In addition to having the capability to version via Git, the move will enable for more contributors and an easier onboarding process. A new contributor can quickly learn Markdown which is simply text with some formatting elements.


For more information, visit the GeneraERS GitHub or Community Page.

Learn more about the GenevaERS and other Open Mainframe Projects in the 2024 Annual Report.

Keep up to date with Open Mainframe Project:

Close Menu