Real-Time Divisible Load Scheduling for Cluster Computing

Real-Time Divisible Load Scheduling with Advance Reservation

“…In this paper, we adopt similar task and system models as our previous work [22,23]. For completeness, we briefly present these below.…”

“…In [22,23], we investigated the problem of real-time divisible load scheduling in clusters. Our previous work, however, does not consider the challenge in supporting advance reservations.…”

“…To specify AvgD, we use the term DCRatio [22]. It is defined as the ratio of mean deadline to mean minimum execution time (cost), that is AvgD E (Avgσ,N ) , where E(Avgσ, N ) [22] is the execution time assuming the task has an average data size Avgσ and is allocated to run on N fully-available nodes simultaneously. Given a DCRatio, the cluster size N and the average data size Avgσ, AvgD is implicitly specified as DCRatio × E(Avgσ, N ).…”

“…As a result, the real-time scheduling of arbitrarily divisible loads is becoming a significant problem for clusterbased research computing facilities like the U.S. CMS Tier-2 sites [33]. Due to the increasing importance [29], a few efforts [16,20,22] have been made in real-time divisible load scheduling, with significant initial progress in important theories and novel approaches.…”

“…To analyze the cluster load for a simulation, we use the metric SystemLoad [22]. It is defined as, SystemLoad = E(Avgσ, N ) × λ, which is the same as, SystemLoad = T otalT askNumber×E (Avgσ,N ) T otalSimulationT ime…”