Skip to main content

Advances, Systems and Applications

Journal of Cloud Computing Cover Image

Table 4 Summary of identified related work classified using the consolidated taxonomy

From: Cloud resource management: towards efficient execution of large-scale scientific applications and workflows on complex infrastructures

Work Highlights MK DL CT DT DY RL SC EN HM WL
[105] Dynamic level scheduling (DLS) x . . o x . . . . x
[126] Wide-area scheduling with dynamic load balancing . . . x x . . . o .
[57] Dynamic Critical Path (DCP) o . . . x . . . . x
[99] Integration to conventional schedulers. . . . . . . . . o .
[2] ELISA, decentralized dynamic algorithm . . . . x . . . o .
[51] Hierarchical scheduling o . . . . . . . . o
[27] Federation of resource traders . . . . . . . . o .
[117] HEFT (Heterogeneous Earliest Finish Time) x . . . . . . . . .
[113] Redundantly distribute job to multiple sites to increase backfilling . . . . . . . . o .
[30] Performance and reliability optimization x . x . x x . . . x
[16] Reduce maximum job waiting time in the queue x . . . . . . . o .
[3] Community of peers for brokering . . . . o . . . o .
[49] Fault-tolerant scheduling . . . . . x . . . x
[79] Dynamic, deadline, energy . x . . x . . x . x
[97] Rescheduling policies x . o . x . . . . x
[58] Auction-based scheduling. . . . . . . . . o .
[138] Deadline partitioning . x o . . . . . . x
[121] Dynamic voltage scaling . x o . o . . x . x
[137] Genetic algorithm to optimize cost with deadline constraint . x x . . . . . . x
[149] Merge multiple DAGs x . . . . . . . . x
[104] Makespan and robustness x . . . . x . . . x
[102] Load balancing on arrival . . . . o . . . o .
[98] LOSS and GAIN approaches x . x . . . . . . x
[43] Performance and reliability optimization x . . . . x . . . x
[31] Reliable HEFT x . . . . x . . . x
[139] Minimize execution time and cost x x x . . . . . . x
[71] Dynamic scheduling . . . . x o . . . x
[90] Dynamic storage mgmt. o . . x . . . . . x
[55] Energy and deadline . x . . o . . x . x
[145] Forecast prototype and SLA compensation . . x . . . . . . .
[146] Historical information, forecasting . . x . . . . . . .
[47] Delegated matchmaking, local vs remote usage . o . . o . . . o .
[29] Improve average response time o . . . . . . . o .
[142] Float time amortization . x o . . . . . . x
[142] Based on HEFT x . . . . . . . . x
[83] Bandwidth speedup, data-intensive o . . x . . . . . x
[89] Makespan and energy x . . . . . . x . x
[133] MQMW (Multiple QoS scheduling of Multi-Workflows) x . x . x . . . . x
[84] RASA (Resource-Aware Scheduling Algorithm x . . . . . . . . .
[59] Decentralized model that improves makespan x . . . . . . . o .
[35] Fuzzy approach for decentralized grids o . . . . . . . o .
[93] Backfilling strategy based on dynamic information x . . . x . . . o .
[23] Ant Colony Optimization x x x . . x . . . x
[140] Path-based deadline partition . x . . . . . . . x
[141] Greedy time-cost distribution . . x . . . . . . x
[61] Optimize makespan and resource utilization x . . . x . . . . x
[114] Similar to YU et al., 2007 x x x . . . . . . x
[13] Data staging x o . x . . . . . x
[96] QoS-aware, cost and execution time x . x . . . . . . .
[153] Based on genetic algorithm; increase resource utilization . . . . x . . . . .
[100] Cost-based . . x . . . . . . .
[67] Time-cost-based, instance-intensive workflows x x x . . . . . . x
[82] Particle swarm optimization heuristic; . . . x x . . . . x
[94] Brokering for multiple grids. . . . . . . . . o .
[122] Bidding system for resource selection o . . . . . . . . .
[128] PSO to minimize cost with deadline constraint o x x . . . . . . x
[39] Optimize makespan and cost x . x . . . . . . x
[88] Dynamic programming x . x . . x . . . x
[25] Dynamic scheduling . . . . x o . . . x
[11] Energy efficiency . . . . o . . x . .
[131] Reputation-based QoS provisioning o o x . . . . . . .
[74] Deadline, budget, auto-scaling . x x . o . . . . .
[64] SHEFT (Scalable HEFT) x . . x x . . . . x
[119] OWS (Optimal Workflow Scheduling); x . . . . . . . . x
[132] Justice-based scheduling x . . . o . . . o .
[151] Budget-constrained HEFT . . x . x . . . . x
[62] CCSH to minimize makespan and cost x . x . . . . . . x
[19] Deadline optimization based on delaying . x . . x . . . . x
[73] Multiple DAGs; deadline-based o x o . x . . . . x
[15] Hybrid clouds; iteratively resch. tasks until mksp.; deadline x x o . x . . . x x
[60] Makespan and energy x . . . . . . x . x
[78] Makespan and energy x . . . o . . x . x
[116] MapReduce on public clouds . x x . . . . . . .
[50] Multi-tier applications o . o . . . . . . .
[143] Auction-based, cloud-provider viewpoint . . . . . . . . . .
[40] Heterogeneous workloads . x . . . . . . . .
[54] SLA management, improve resource utilization . o o . . . . . . o
[107] Multi-cloud, cost optimization . . x . . . . . x .
[37] Multi-objective, cost constraints . . x . . . . . x .
[144] Backtracking and continuous cost evaluation o . x . x . . . . x
[33] Multi-objective scheduling x . x o . x . x . x
[12] Pareto-based; execution time and cost x . x . . . . . . x
[118] Combination of DAG merging techniques x . . . . . . . . x
[70] Auto-scaling of resources o x x o . . . . . x
[124] Fault-tolerant scheduling . x x . o x . . . x
[120] Deadline-driven, scientific applications, hybrid clouds . x . . . . . . x o
[148] Energy-aware, scheduling delay . o . . o . . x . .
[20] Aneka platform; QoS-driven, hybrid . x . . o . . . x .
[48] Cost minimization, deadline . x x . . . . . . .
[26] Negotiation/bargaining . x x . . . . . . .
[129] Market oriented x . x . . . . . . x
[46] Community-aware decentralized dynamic scheduling o . . . o . . . o .
[1] Partial Critical Path (PCP) o x o . . . . . . x
[65] Minimize end-to-end delay o . x . . . . . . x
[152] Monte Carlo approach x . o . x . . . . x
[103] Power aware scheduling x . . . o . . x . x
[134] Particle swarm optimization x . x . o . . x . x
[130] Data-intensive, energy-aware x . . x o . . x . x
[42] Rule-based . o o . . . . . x .
[38] Energy, deadline . x o . . . . x . .
[41] Bag of tasks, time and cost x . x . . . . . . .
[106] SLA-based cost model; power o . x . . . . o . .
[136] Cost management . . o . . . . . . .
[5] Predict Earliest Finish Time (PEFT) x . . o . . . . . x
[14] Cat Swarm Optimization . . x x . . . . . x
[95] PSO considering performance variation and VM boot time . x x . x . . . . x
[4] Aggregation-based budget distribution . . x . . . . . . x
[87] Critical-path heuristic x x x . . x . . . x
[86] Spot instances . x x . x x . . . x
[10] Fault-tolerance . . x . o x . . . .
[56] Behavioral-based estimation . . o . x . . . . .
[154] Multiple workflows, optimize time and cost x x x . . . . . . x
[68] Multi-cloud, enhanced workflow model o x x x . . . . x x
  1. MK = Makespan/Time; DL = Deadline; CT = Cost/Budget; DT = Data-Intensive; DY = Dynamic; RL = Reliability; SC = Security; EN = Energy; HM = Hybrid/Multicloud; WL = Workload/Workflow;. = Not addressed; x = Fully addressed; o = Partially addressed