“…Best-effort based scheduling attempts to minimise the execution time without considering other factors such as the monetary cost of accessing resources and various users' QoS satisfaction levels. Some examples are the Heterogeneous EarliestFinish-Time algorithm (Tannenbaum, Wright, Miller and Livny 2002) used by ASKALON (Fahringer, Jugravu , Pllana , Prodan, Slovis and Truong 2005), the Min-Min algorithm (Maheswaran, Ali , Siegel , Hensgen and Freund 1999) used by GrADS (Berman, Chien, Cooper, Dongarra, Foster, Gannon, Johnsson, Kennedy, Kesselman, MellorCrummey, Reed, Torczon and Wolski 2001) and a throughput maximisation strategy used by SwinDeW-G (Yang, Liu, Chen, Lignier and Jin 2007) and SwinDeW-C (Yang, Liu, Chen, Liu, Yuan and Jin 2008). In contrast, QoS constraint based scheduling attempts to maximise the performance under QoS constraints of which cost is one major constraint that we focus on in this paper.…”