August 2023: Leveraging Adaptive Framework for Open Source Data Access Solutions

More than a decade ago, Clemson University outlined the requirements needed to integrate several campus-wide enterprise applications in a way that would automate the exchange of data between them, and establish the relationships of that data to the unique identities that represented all users within the system, including faculty, staff, students, alumni and applicants. There would be no direct access of data, except through applications that were approved and had established Memorandum of Understanding (MOU) contracts in place. This project was known as the Clemson Vault.

Within the Identity Management space, solutions for automating the provisioning of identities are offered by several vendors these days. However, mileage and cost vary when you wish to integrate arbitrary university resources, such as mailing lists, disk storage, building card access, and course registrations. Open source solutions, with all of the above requirements, are non-existent.

At Clemson University, we combined licensed vendor software and in-house apps, scripts and procedures to create a data integration solution that met the original requirements. This implementation has served us well for many years, but many of the drawbacks to the current design prompted us to begin pulling out many of these features into its own project, where we could collaborate on features and enhancements for the future with institutions outside of our own organization. The patterns, interfaces, and source code that emerged from the original vault were extracted out, embellished and migrated into an open source repository known as Adaptive Framework (https://github.com/afw-org/afw).

Clemson University has been working on this project for several years now, and has recently released this open source framework for building data access solutions that provide web service API’s, data transformation tools, real-time data provisioning and an authorization architecture. The framework that has emerged offers a built-in scripting language, pre-compiled server-side applications and an administrative web interface.

Although it was originally designed for the implementation of an open source identity vault, we envision a broader adoption of this framework for other data-driven needs, such as extending databases with metadata, building policy-based authorization systems, and integrating data repositories with a metadata catalog, and varying levels of access control, across federated environments.

Our goal with this project is to gather external support from both commercial and public institutions to help make this framework sustainable moving forward.

Speaker Bio:

Jeremy Grieshop is a software engineer (B.S. Miami University, M.S. Clemson University) and has been employed by Clemson University since 2001. His role has been in software development for the Identity Management team and has been directly involved in the software design and implementation of many of the authentication and provisioning software, along with self service tools that are in place at Clemson University today.

Jeannette Dopheide