| As an important branch in the field of Blockchain technology,the Consortium Blockchain has become a research hotspot in the industry.At present,Hyperledger Fabric project incubated by the Linux Foundation has become a benchmark in the global Consortium Blockchain field.However,the Fabric system has the problem of tight coupling of computing and storage similar to the traditional distributed database architecture.Enterprises lack flexibility when deploying node clusters,node resources cannot be fully utilized,and it is difficult to achieve rapid and elastic expansion of cluster resources,which ultimately leads to the failure of system transaction efficiency to meet expectations.This not only increases the burden of enterprise operation and maintenance,but also raises the participation threshold of the Consortium Blockchain.Based on the research and analysis of the theory of separation of Computing-Storage Separation and the existing mature projects applying this theory,this paper proposes a Computing-Storage Separation Scheme that effectively adapts to Fabric according to the characteristics of the Fabric system.The main contributions of this paper are as follows:(1)A new distributed computing and storage separation model that effectively adapts to the Fabric system is designed to realize the decoupling of the computing and storage resources of the Fabric system.(2)Design and implement a distributed cache mechanism to relieve the performance constraints of computing and storage resources.The cache mechanism meets the data ordering and reliability required by the Fabric system on the basis of ensuring data availability and eventual consistency.At the same time,the distributed structure design In line with the concept of Blockchain decentralization.(3)Solve the problems of computing and storage cluster communication,resource expansion and recovery,and multi-ledger constraints.Cluster communication ensures the consistency of data within the cluster and the overall data consistency of the system.Resource expansion and recovery include dynamic node expansion and contraction and rapid regeneration.In addition,the multi-ledger isolation of computing and storage clusters is implemented to maintain the diversity of the Fabric system ledger.The main contents of this paper are:(1)Design Fabric-Based Computing-Storage Separation Model.(2)Detailed Design and Implementation of Fabric-Based ComputingStorage Separation System.(3)Testing of Fabric-Based Computing-Storage Separation System.The final test results show that the computing and storage separation scheme designed and implemented in this paper has low coupling between computing and storage,and the performance is not restricted by each other.It has the ability to dynamically exchange and expand computing and storage resources while maintaining the same performance as the original architecture.The native architecture shows higher deployment flexibility and fuller resource utilization,and the node recovery time is greatly reduced. |
