Application Data Collection Toolkit
Application Data Collection Tools (ADCT) is a scalable software tool set for gathering, transporting, storing, and retrieving low-volume messages containing time-sensitive data from computational modeling and simulation workflows on clouds, servers, desktops and HPC clusters.
- It enables the collection and reuse of machine-readable data for machine-learning and artificial intelligence based research in the areas of computing performance modeling, computational system management.
- It enables the collection of feature-level software use information for organizational understanding of formal or informal computational workflows.
- It is designed for use across complete computing workflows.
- It is implemented to support very long data life-times and based on open standards (JSON, HTTPS).
- It is licensed to support an open-development model.
- It is, from an application-development viewpoint, a scalable, simple structured logging library.
This is the umbrella repository for the integration, testing, and release of various portions of ADC tooling.
Associated code repositories are (or will soon be):
- https://github.com/sandialabs/adct-json
- https://github.com/sandialabs/adct-cxx
- https://github.com/sandialabs/adct-python
- https://github.com/sandialabs/adct-java
Additional repositories are planned for documentation, REST APIs, server implementations, and related data management tools and examples. Additional language bindings will be included when available from the development community.