The R Esearch Of Live Virtual Machine Migration Based On HashGraph

With the rapid development of the cloud computing and the continuous expansion of data center clusters,the connection between data centers is becoming closer and closer.Live virtual machine(VM)migration technology has become an effective method to solve hot issues such as load balancing and disaster recovery.Live VMs migration across data centers requires to transfer large amounts of data,but there is a problem of unstable network latency in the WAN environment.How to solve the problems of large data transfer and long migration time during the live migration across data centers has become a research hotspot and difficult point in cloud computing.This article focuses on two methods in the current field of live VMs migration across data centers.One is to directly migration between the two data centers;the other is to provide global shared storage system between different data centers to store VM images.For reducing the amount of data transferred,both methods deduplicate VM images during migration.However,the deduplication is a computation-intensive operation,which will affect the migration performance of the VMs,and the global shared storage system is a third-party,which has low security.Combining the advantages and disadvantages of the existing methods,We proposed an improved method of live VM migration based on hashgraph.The main work of this paper is as follows:(1).Discussion and analysis of work related to live VM migration.This paper introduces the development of live VM migration,and summarizes related work in the field of VM migration from three aspects:memory data migration,storage data migration,and network connection status.We deeply study of existing live VM migration methods across data center,and analyze the advantages and disadvantages.(2).The paper proposes a method of live VM migration across data centers based on hashgraph.Firstly,we introduce a migration agent for the data center to improve management capabilities for live migration across data centers.Secondly,we use the migration agent to build a hashgraph network.,it can reach a consensus on the hash of images through the gossip protocol and construct a storage of image fingers based on Merkle DAG to optimize the efficiency of deduplication.(3).Experimental evaluation and analysis.Firstly,we summarize the performance indicators of live migration,and then analyze the effectiveness of the method of live migration across data centers based on hashgraph.By analyzing the simulation experimental data,compared with the existing method,the total migration time of the method in this,paper under the four different similarities is reduced by 39.3%,37.9%,43.7%,43.5%on average,and the total network traffic is also reduced.Lastly,summary the paper and prospect the future.