General Information
The Fast INtention Agnostic LEarning Server (FINALES) is a message broker for coordinating communication in big scientific projects.
Through the endpoints of the FINALES REST-API, the tenants (implemented as clients) can send requests for data or materials, view the requests currently pending in the server, and post the results requested by others. Each tenant is "intention agnostic", in the sense that it doesn't need to know the broader context of a given request. It will decide to service it or not depending on the individual specifications alone (can it produce that type of measurement? can it go to the given temperature? can the requester afford the cost of the measurement?).
Below you will find the basic installation instructions of the package. Please check the rest of the documentation in the wiki of the project for instructions on how to:
- Manage a FINALES server (for admins)
- Interact with a FINALES server (for tenants).
The server release 1.0.0 employs a comprehensive relational database design shown below
By utilising uuid's, timestamps and logging tables the server employs an (almost) append only structure.
For the append only structure to be enforced is_active columns in the request and results table should be handled by logging tables, which is planned for the next release(small changes to the status columns are also needed). The (almost) append only design and timestamps allow for easy traceability and reproducibility in a server campaign.
As shown in the figure above each tenant, request and result will be associated with a uuid. These uuid's are employed as primary keys and map tables to eachother through foreign keys. These keys will identify any tenant, request and result in the server, which will be used as the id of specific objects in a final ontology mapping.
Code that have specifically been written by DTU employees is a copyright © 2023 Technical University of Denmark.