A QoS-aware Model for Cloud Storage Environment

Nowadays, the administration of resources and application scheduling in a large-scale cloud storage environment is a formidable task, especially in dynamic changing scenario both in storage providers and consumers requirement sides. In order to meet the challenge of dynamic change characteristic in cloud storage application, this paper presents a new QoS-aware approach to match individual applications to most appropriate cloud storage nodes based on their particular attribute matrix, such as capacity, response time, I/O throughput, availability and user requirements etc. The experimental results illustrate that our QoS-aware method can achieve better performance than existing random scheduling algorithm in terms of load balance, response time and I/O throughput.