Research On Distributed Storage Management Of Educational Resources Based On Value Evaluation Strategy

In the era of education informatization 2.0,digital education resources have exploded,and traditional stand-alone file systems have been unable to accommodate these massive educational resources.HDFS(Hadoop Distributed File System)is a leader in distributed storage systems.Installed on a low-cost machine,it has the characteristics of high fault tolerance,high reliability and streaming access,which is very suitable for storing massive educational resources.Although HDFS is very advantageous as a storage system for massive educational resources,there are areas where native HDFS can be improved:On the one hand,as the storage of educational resources increases,the default placement algorithm used by HDFS causes the uneven load of data nodes.Obviously,this will seriously affect the performance of the file system;on the other hand,HDFS uses a multi-copy redundancy strategy for all educational resources without distinction to ensure data reliability,which will bring a huge storage cost to the file system.In order to improve the shortcomings of the HDFS file system,this paper finds that the current research mostly starts with the heat of files and the heterogeneity of cluster nodes.The file heat is also the main reference index to guide algorithm improvement.For the storage of educational resources,the popularity of files is the popularity of educational resources.Compared with other file resources,whether educational resources are popular with learners is closely related to the value of their resources.Therefore,this article will study the improvement of HDFS placement algorithm and data redundancy strategy based on the value of educational resources.First of all,in order to realize the evaluation of the popularity of educational resources,this paper establishes an evaluation index system of the value of educational resources.When selecting the evaluation indexes,it refers to the general characteristics of popular educational resources and various specific educational resource evaluation systems constructed by previous people;The establishment of an educational resource evaluation system is to guide the improvement of subsequent placement algorithms and resource redundant storage strategies.Therefore,educational resources are graded by constructing a value rating model of educational resources.Then,using the divided educational resource levels,this paper designs a storage placement algorithm based on educational resource ratings.This algorithm additionally records the data volume of each rated educational resource in each data node.When the resource is stored,it will integrate each The storage data amount of the data node and the storage data amount of the rating corresponding to the stored resource are sorted to select a most suitable storage node.Finally,in view of the fact that traditional multiple copy storage strategies consume a lot of resource space,this paper proposes a data redundancy storage strategy based on educational resource ratings.According to the different popularity of educational resources with different ratings,it is used for educational resources with high ratings.Multi-copy redundant storage scheme to ensure the reliability of resources while improving system load capacity,adopt erasure coding technology for redundant storage of educational resources with lower ratings,and propose a dynamic conversion algorithm to facilitate two redundant storage schemes Conversion.In this paper,in order to verify the effectiveness of the algorithm,simulation experiments are carried out in the corresponding chapters that propose relevant improved algorithms or strategies.Through comparison and analysis with the experimental results of HDFS default algorithms or strategies,it is concluded that the improved placement algorithm implements each node Storage load balancing,improved access load capacity,and improved storage strategies reduce space consumption and increase storage utilization.