Prepartition:A New Paradigm for Load balance in Cloud Data Centers

It is significant to apply load balancing strategy to improve the performance and reliability of resource in Cloud data centers. In a dynamic changing traffic environment, workload of a Cloud data center has real-time and fixed process time characteristics. One of the challenging scheduling problems in Cloud data centers is to take the real-time allocation and migration of reconfigurable virtual machines as well as the integrated features of hosting physical machines into consideration. In general, load-balance scheduling is NP-hard problem as proved in many open literatures. Traditionally, for offline load balance without migration, one of best approaches is LPT (Longest Process Time first), which is well known to have approximation ratio 4/3. With virtualization, reactive post migration of virtual machines (VMs) after allocation is one populer way for load balance and traffic consolidation. However, reactive migration may cause interruption and instability of service and other associated costs. In view of this, we propose a new paradigm: preactive migration by prepartition of virtual machines to set process-time bound for each request on each physical machine (PM). Prepartition can reduce process time by preparing VM migration in advance therefore reduce instability and achieve better load balance. Synthetic and trace-driven simulation results show that Prepartition has better performance than the best known load balancing algorithms with regard to total imbalance level, makespan, overall load efficiency as well as capacity makespan.