Resource Federation
A European Compute and Data Continuum must allow resources from different providers and domains to function as part of a single, coherent ecosystem. For HEP and RA, where workflows routinely span multiple facilities, this means going beyond isolated access models to enable true federation of access and allocation of continuum resources. Researchers should be able to use HPC, cloud, storage, and edge resources through a unified entry point, with allocations that can be honoured seamlessly across institutional and national boundaries.
At the technical level, this requires interoperable standards for tasks like job submission, data access, and monitoring, as well as the integration of federated identity and access management systems. Building on existing models such as eduGAIN and the WLCG’s Virtual Organisations, the continuum should provide single sign-on across facilities, with support for role-based access, attribute aggregation, and support for service accounts. These mechanisms are essential for ensuring both ease of use and robust security in a federated environment.
Transparent accounting is equally important. Usage data, such as CPU/GPU hours, storage consumption, and data transfer volumes, must be collected and reported in a consistent way across all sites. This not only supports fair allocation and equitable participation but also enables funding bodies and infrastructure providers to verify contributions and assess impact. For researchers, transparent accounting builds confidence that resources are distributed fairly, while for policymakers, it provides the accountability needed to sustain investment.
For HEP, this would ensure that exabyte-scale workloads are distributed efficiently across a federation of Tier-0, Tier-1, and Tier-2 centers, with harmonised allocation and usage reporting. For RA, it would allow data from telescopes to flow into European HPC and HTC facilities in a controlled and federated manner, providing astronomers with seamless access to the resources they need for continuous, multi-decade observations.