Enterprise Data Center Disaster Recovery System Architecture Design And Implementation

As IT systems are increasingly moving toward centralization of data and services, the data center must provide resistance to vandalism and the ability to respond to natural disaster order to reduce the enormous impact and business losses caused by system downtime when IT systems encounter a variety of unexpected failures and disasters. Based on mainstream disaster recovery system construction technologies, combined with the actual business needs of an enterprise, this thesis designs and implements a highly available enterprise disaster recovery system. The specific tasks of this thesis are to:(1) Analyze enterprise disaster recovery system construction requirements, and based on the business characteristics of the enterprise, develop RTO (Recovery Time Objective) and RPO (Recovery Point Objective) indicators; give a detailed analysis of the shortage of the current disaster recovery construction, and design a disaster recovery system framework needed to fulfill the indicators based on business characteristics and disaster recovery needs.(2) Design and implement a multiple-outlet network topology for disaster recovery system based on BGP and OSPF network protocols, use OSPF protocol for routing and switching of Intranet among three data centers, and complete controls of routings by route-map so that the three centers form a unified Intranet structure. Use BGP protocol for Internet access, leverage BGP routing broadcast properties to announce anycast routing in the three data centers which can be regarded as three different exits of the entire enterprise network, independent of each other and standby for each other, thus guaranteeing high availability of the corporate extranet.(3) Design and implement the solution for replicating data among three centers in two locations for the disaster recovery system based on storage replication technology, use SRDF synchronous replication technology for synchronous data replication between the intra-city data center and the primary data center, and each time data is written to the primary data center and the intra-city data center at the same time before returning to users, thus guaranteeing zero data loss. Use SRDF asynchronous replication technology for asynchronous data replication between the offsite disaster recovery center and the primary data center. When large-scale irresistible disasters occur, the offsite data center can quickly take over services, and data loss is kept in the minimum tolerable range.(4) Validate the operation of the "Three Centers in Two Locations" disaster recovery system architecture. Perform routing tests from different locations and validate the effectiveness of the disaster recovery center in taking over services when disaster recovery occurs.While minimizing the company’s overall investment cost, the "Three Centers in Two Locations" disaster recovery system construction solution makes up for the shortcomings in enterprise disaster recovery system construction and enhances business continuity. When disaster occurs, it can restore service as quickly as possible without loss of data, keeping business losses to a minimum.