Towards a Production Volunteer Computing Infrastructure for HEP

Abstract

Abstract. Following the successful inclusion of virtualisation to volunteer computing for theory simulations back in 2011, the use of volunteer computing with BOINC and CernVM has been extended to cover simulations for the LHC experiments ATLAS, CMS and LHCb. This paper describes the status of the BOINC volunteer computing platform at CERN used for LHC@home and how it has been designed to address a heterogeneous environment of different user communities with different computing infrastructure. The aim of the recent developments is to provide a volunteer computing platform that the experiments can build upon to exploit opportunistic resources. Furthermore, new developments of common solutions to span user authentication domains are explained.

Browser-based Harnessing of Voluntary Computational Power

“…It was suggested that lottery with monetary awards could be used to motivate users [68]. Awards could be printed diplomas diplomas for top contributors as in Einstein@Home [36]. Users could get credits which could be then used to pay for accessing web resources [29] [64].…”

“…Users could get credits which could be then used to pay for accessing web resources [29] [64]. However, even purely virtual credit system providing nothing more except the reputation can be very successful to attract and keep the users [3] [36]. To increase the reward for users, credits should be assigned as soon as possible, with publicly visible rankings [3].…”

“…To decrease the amount of effort in developing new BOINC applications and their maintenance, a virtual machine approach was implemented [36]. First, the application is compiled for a particular virtual machine.…”

LHC@Home: a BOINC-based volunteer computing infrastructure for physics studies at CERN

“…Following the experience with Test4Theory, ATLAS@Home and other pilot projects, with a view to include volunteer computing into the production computing infrastructure for HEP [12], a major e ort has been undertaken to consolidate the original LHC@Home and host additional applications utilising virtualisation.…”

Explore multi core virtualization on the ATLAS@home project

“…With this successful use case, people started to explore the possibility of running the platform dependent HEP experiment software on volunteer computers. This required virtualization technologies which would convert the heterogeneous volunteer computers into the "volunteer clouds" [6] and provide the suitable operating system and software environment that all the HEP experiment computing require.The [12]started to explore the possibility of harnessing volunteer computing resources and integrating them into their grid computing platforms, as the grid computing platforms are usually the major computing resources for HEP computing and upon which the computing workflows are designed. ATLAS was the pioneer in this field with the success in both gaining a wide range of publicity and the amount of free resource from its volunteer computing project ATLAS@home [13][14].…”