Mial, A Virtual Resource Allocation Algorithm In The Application Of AutoLab

With the rapid development of cloud computing and virtualization, the management of virtual allocation of resources has become a particularly important topic, and virtual resource allocation algorithm is also growing in importance. Good resource allocation algorithm can make the overall system efficiency, play a multiplier effect. Major cloud computing companies such as Google, Amazon, Microsoft and IBM have their own more mature virtual resource allocation algorithms, these algorithms have been successfully applied to their cloud computing products.The AutoLab project is an internal application of virtual resource distribution management software. According to requirements of the tasks submitted by the user it makes a rational allocation of system resources. The original algorithm in AutoLab is SIAL (Single In AutoLab), the idea of the algorithm is clear and straightforward, its relatively simple algorithm logic can barely fulfill system goals. The SIAL sometimes makes misjudgment, the tasks which could be assigned are judged to be impossible. SIAL even may frequently cause load imbalance.In order to resolve the remaining issues in the SIAL, this paper focuses on analysis and virtual resource allocation algorithms with major cloud computing companies such as Google, Amazon, Microsoft and IBM. We design and implement MIAL (Multiple In AutoLab) according to characteristics of AutoLab. MIAL splits the task into pieces and assigns them to multiple servers instead of distributing the whole task on the same virtual machine. In this way, MIAL enhances the load balance of AutoLab. MIAL evolves with the evolution of AutoLab, its architecture is becoming more and more clear with good scalability.MIAL reduces the misjudgment caused by SIAL and improves the system load balance. The pressure of one single server is also reduced. Overall MIAL can well meet the requirements of AutoLab and solve the issues left by SIAL. However there will be a few load imbalance and misjudgment occasionally, as a result MIAL needs further improvement in the future.