A Scheduling Algorithm for Defeating Collusion

By exploiting idle time on volunteer machines, desktop grids provide a way to execute large sets of tasks with negligible maintenance and low cost. Although desktop grids are attractive for cost-conscious projects, relying on external resources may compromise the correctness of application execution due to the well-known unreliability of nodes. In this paper, we consider the most challenging threat model: organized groups of cheaters that may collude to produce incorrect results. By using a previously described on-line algorithm for detecting collusion and characterizing the participant behaviors, we propose a scheduling algorithm that tackles collusion. Using several real-life traces, we show that our approach minimizes redundancy while maximizing the number of correctly certified results. Key-words: Volunteer computing systems, collusion, scheduling, result certification, BOINC ∗ Nancy University, LORIA and LaBRI – Nancy & Bordeaux, France † INRIA and LaBRI – Bordeaux, France ‡ Dept. of Computer Science and Engineering – University of Minnesota, Twin Cities – Minneapolis, USA in ria -0 05 24 49 3, v er si on 1 8 O ct 2 01 0 Un algorithme d’ordonnancement pour déjouer la collusion Résumé : Ce rapport présente un algorithme de placement de tâches dans les systèmes de calcul volontaires permettant de déjouer la collusion (attaque coordonnée des travailleurs volontaires). Il se base sur nos précédents travaux sur la détection de collusion. Les résultats expérimentaux utilisant des traces concrètes montrent que notre approche minimise la redondance et maximise le nombre de résultats certifiés correctement. Mots-clés : Systèmes de calcul volontaires, collusion, ordonnancement, certification de résultats, BOINC in ria -0 05 24 49 3, v er si on 1 8 O ct 2 01 0 Defeating Collusion 3