Answers are grounded in this paper and cite their sources.
Cloud storage services keep user data on many storage nodes, and two decisions decide how well they perform: which node serves a request, and how many copies of the data are kept. Common policies such as First-Come-First-Served (FCFS) and Round Robin ignore node capacity, load and bandwidth, and most file systems apply one fixed replication factor to all data. This paper presents a cloud data storage scheduling framework that takes both decisions together. An efficient scheduler gives every feasible node a composite score from its free capacity, load, bandwidth and latency and picks the best one. A replication manager then derives the replication factor from a target availability, the node failure probability, and the access frequency and criticality of the object, and places the replicas on distinct nodes with automatic failover. The environment is modelled in CloudSim and Storage CloudSim, and the scheduler will be compared with FCFS, Round Robin, Shortest Job First and Min-Min on makespan, response time, throughput, load imbalance, availability and storage cost. An administrator dashboard shows node load, replica status and alerts. The work is at the design stage, so the paper gives the architecture, the mathematical model and the evaluation plan, and no experimental results; the worked example is illustrative.
No takes yet. Share an insight, caveat, or question.
Patare et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: