Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Federated Resource Allocation for HPC Services

Livenson, Ilja ; Saar, Ahti ; Adomeit, Marina ; Lysø, Eivind ; Hansson, Emma Lisa LU ; Laitinen, Jarno and Sjöström, Anders LU (2025) 6th Nordic e-Infrastructure Collaboration Conference, NeIC 2024 In Communications in Computer and Information Science 2398 CCIS. p.128-141
Abstract

This paper presents an implementation of a federated resource allocation system (Puhuri), which offers a full set of functions from the application phase for computing infrastructure resources to consumption reporting. The implementation is based on the Puhuri project of Nordic e-Infrastructure Collaboration (NeIC) [1]. Shared research infrastructures, particularly high-performance computers, necessitate a robust mechanism for allocating resources in accordance with predefined policies. An allocation serves as authorisation to utilise a system in an agreed-upon manner. The impact of Puhuri on shared research infrastructures is to enable multiple resource allocation teams to allocate resources across different domains and organisations.... (More)

This paper presents an implementation of a federated resource allocation system (Puhuri), which offers a full set of functions from the application phase for computing infrastructure resources to consumption reporting. The implementation is based on the Puhuri project of Nordic e-Infrastructure Collaboration (NeIC) [1]. Shared research infrastructures, particularly high-performance computers, necessitate a robust mechanism for allocating resources in accordance with predefined policies. An allocation serves as authorisation to utilise a system in an agreed-upon manner. The impact of Puhuri on shared research infrastructures is to enable multiple resource allocation teams to allocate resources across different domains and organisations. Puhuri has integrated the LUMI EuroHPC supercomputer [2] and the biodiversity digital twin workflow management framework Lexis. Additionally, Puhuri has also been collecting requirements and doing technical integrations in a wider scope. Users accessing the system must be unequivocally identified. This process, known as authentication, ensures that the right individuals utilise the allocated resources to which they are authorised by the resource allocation team. GÉANT has established the MyAccessID service [3] for federated authentication, which Puhuri also utilises. One of the challenges is related to the user identity vetting for the users whose identity provider is not signalling that they are verifying users’ identities in the required manner, which requires them to use commercial third-party services for automatic identity vetting based on user’s identity cards.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; and
organization
publishing date
type
Chapter in Book/Report/Conference proceeding
publication status
published
subject
keywords
Accounting, Authentication, Authorisation, Federation, Governance, HPC, Identity vetting, Resource allocation
host publication
Nordic e-Infrastructure Tomorrow - 6th Nordic e-Infrastructure Collaboration Conference, NeIC 2024, Proceedings
series title
Communications in Computer and Information Science
editor
Azab, Abdulrahman and Malkiewicz, Tomasz
volume
2398 CCIS
pages
14 pages
publisher
Springer Science and Business Media B.V.
conference name
6th Nordic e-Infrastructure Collaboration Conference, NeIC 2024
conference location
Tallinn, Estonia
conference dates
2024-05-27 - 2024-05-29
external identifiers
  • scopus:105002558326
ISSN
1865-0937
1865-0929
ISBN
9783031862397
DOI
10.1007/978-3-031-86240-3_9
language
English
LU publication?
yes
id
c0a78cf3-eddb-4336-bd00-b291698f93d1
date added to LUP
2025-08-29 13:43:06
date last changed
2025-08-29 14:20:37
@inproceedings{c0a78cf3-eddb-4336-bd00-b291698f93d1,
  abstract     = {{<p>This paper presents an implementation of a federated resource allocation system (Puhuri), which offers a full set of functions from the application phase for computing infrastructure resources to consumption reporting. The implementation is based on the Puhuri project of Nordic e-Infrastructure Collaboration (NeIC) [1]. Shared research infrastructures, particularly high-performance computers, necessitate a robust mechanism for allocating resources in accordance with predefined policies. An allocation serves as authorisation to utilise a system in an agreed-upon manner. The impact of Puhuri on shared research infrastructures is to enable multiple resource allocation teams to allocate resources across different domains and organisations. Puhuri has integrated the LUMI EuroHPC supercomputer [2] and the biodiversity digital twin workflow management framework Lexis. Additionally, Puhuri has also been collecting requirements and doing technical integrations in a wider scope. Users accessing the system must be unequivocally identified. This process, known as authentication, ensures that the right individuals utilise the allocated resources to which they are authorised by the resource allocation team. GÉANT has established the MyAccessID service [3] for federated authentication, which Puhuri also utilises. One of the challenges is related to the user identity vetting for the users whose identity provider is not signalling that they are verifying users’ identities in the required manner, which requires them to use commercial third-party services for automatic identity vetting based on user’s identity cards.</p>}},
  author       = {{Livenson, Ilja and Saar, Ahti and Adomeit, Marina and Lysø, Eivind and Hansson, Emma Lisa and Laitinen, Jarno and Sjöström, Anders}},
  booktitle    = {{Nordic e-Infrastructure Tomorrow - 6th Nordic e-Infrastructure Collaboration Conference, NeIC 2024, Proceedings}},
  editor       = {{Azab, Abdulrahman and Malkiewicz, Tomasz}},
  isbn         = {{9783031862397}},
  issn         = {{1865-0937}},
  keywords     = {{Accounting; Authentication; Authorisation; Federation; Governance; HPC; Identity vetting; Resource allocation}},
  language     = {{eng}},
  pages        = {{128--141}},
  publisher    = {{Springer Science and Business Media B.V.}},
  series       = {{Communications in Computer and Information Science}},
  title        = {{Federated Resource Allocation for HPC Services}},
  url          = {{http://dx.doi.org/10.1007/978-3-031-86240-3_9}},
  doi          = {{10.1007/978-3-031-86240-3_9}},
  volume       = {{2398 CCIS}},
  year         = {{2025}},
}