Research On Task Scheduling Strategy For Cloud Storage

With the importance rising of cloud storage in the data storage, the task scheduling strategy of cloud storage has become a very important aspect. The task scheduling policy of cloud storage mainly researched the process of cloud storage tasks distribution and completion. With the development of science and technology, cloud storage task scheduling strategy has become a key part of the cloud storage.Cloud storage task scheduling strategy is mainly to sol’-the problem of how to choose a cloud storage system and the better task scheduling(?) to complete the jobs submitted by users.This paper researched task scheduling policy based on three major distributed file systems architecture that include in GFS, HDFS and Dynamo. The main influencing factors of Cloud scheduling strategy include that a copy of the coverage indicators, the cost of maintaining replica consistency, the availability of a copy of the virtual node load, virtual node overhead, network transmission reliability, data replica consistency, available network bandwidth, network transmission delays. The four kinds of user are divided by the main factors, according to the different attention. The users include that individual users, growing corporate users, the general scale enterprise users, corporate users of financial.Combined with AHP level analysis to determine nine factors of Cloud storage task scheduling strategy for each of the four kinds of user weights. Weights propose Cloud storage task scheduling strategy based on the user’s perspective. Finally, optorsim module of the simulation software was extent in the distributed system.Through The modules expansion in distributed simulation software system optorsim carried out a policy simulation.The integrated simulated data concluded:1) Financial enterprise users consider many factors, accounting for more weight in the entire decision-making.2) The growth enterprise users and the general scale enterprise users are similar to the values of the right to share in decision-making.3) Fewer individual users consider the factors accounted for less weight in the entire decision-making.4) In the four kinds of user submitted task queue simulation draw each node suitable for the user. And drawn based on the user’s perspective, cloud storage task scheduling strategy in task execution time is slightly longer, but with better availability and consistency.