Introduction to three centers in two places for disaster recovery and backup
The "three centers in two places" in the disaster recovery solution of two places and three centers generally refers to oneproduction center and one a>. to an off-site disaster backup centerasynchronously to the disaster backup center in the same city. At the same time, the data of the production center is copied synchronouslyThe data of the production center. offsite disaster backup center, and oneIn-city disaster backup center
The local disaster recovery center usually has the same business processing capabilities as the production center. Applications can be switched to the local disaster recovery center without losing data to maintain continuous business operation. When the production center and the disaster recovery center in the same city are unavailable at the same time, business recovery can be achieved in the remote disaster recovery center to maintain business continuity.
Compared with only establishing a disaster backup center in the same city or a remote disaster backup center, the "two places and three centers" approach combines the advantages of the two and can adapt to a wider range of disaster scenarios. For small-scale regional disasters and larger-scale natural disasters, Disasters can be responded to quickly through the disaster backup system, ensuring that business data is not lost as much as possible, and achieving better RPO and RTO. Therefore, the two-place, three-center disaster recovery solution has been widely used.
Schematic diagram of two places and three centers
The disaster recovery method of two places and three centers is an important trend in current disaster recovery construction and is widely used in fields such as telecommunications, finance, manufacturing, and university informatization. Achieve data protection close to zero data loss by building a close data center (data center in the same city), and obtain long-distance data protection by building a data center farther away (offsite data center) , to avoid regional disasters that lead to unrecoverable business.
Common solutions for three centers in two places include the following networking solutions:
Cascade networking
Parallel networking
Comparison of the advantages and disadvantages of the two options
Cascading scheme:
Advantages: Small consumption of local production-end resources and low network bandwidth requirements for remote data transmission;
Disadvantages:When the local production center and the local disaster recovery center fail at the same time, the remote disaster recovery computer room cannot take over the business in real time.
Parallel scheme:
Advantages: The business is highly real-time. When the local production center and the local disaster recovery center fail at the same time, the remote disaster recovery center can take over the business in real time;
Disadvantages:The local production center needs to synchronize data to the local disaster recovery center and the remote disaster recovery center in real time at the same time, which puts great pressure on the local production end and requires bandwidth for remote transmission. higher.
Key component technical implementation instructions
MAN requirements:
Disaster recovery network distance: <100km, bare fiber connection.
Transmission delay: <1ms (one-way).
Real network bandwidth: > Peak write IO bandwidth of the business.
WAN requirements:
Disaster recovery network distance: unlimited.
Transmission delay: <50ms (one-way).
Real network bandwidth: >Average write IO bandwidth of the business.
Disaster recovery management control terminal:
Management workstations require communication between the three centers.
Network distance requirement: No limit.
Communication network bandwidth requirements: 10Mb/s.
Key technical principles of three centers in two places
Application-level high-availability business takeover
Provides high-availability services at any distance for a variety of applications. When application exceptions or production system abnormalities (such as abnormal service stop, network exception, hardware failure, production system downtime maintenance) cause the application business system to be unavailable, the relevant The application is immediately switched to the disaster recovery server, and the application on the disaster recovery server provides services to ensure the continuity and uninterruption of the overall business.
Key technical principles of three centers in two places
Data real-time/scheduled backup
Continuous data backup and on-demand recovery services can easily back up production data in real time or regularly to local or remote disaster recovery center nodes, and quickly restore required data on demand. And strictly ensure the consistency and integrity of data in the production system and disaster recovery center. It can be widely used in real-time disaster recovery backup protection of ordinary file systems, database systems, email systems, etc.
While data is being backed up, the changed data will be copied to the disaster recovery center in real time using byte-level copy or snapshot, and the data changes will be recorded in the form of logs. When the system fails, according to the data change log, the time point that needs to be restored is quickly located, and the data can be restored to the abnormal point with one click to ensure data security and business continuity.
Program Highlights
Byte-level incremental data capture: Using bytes as the smallest unit of data capture, rather than the traditional file or block unit, thus greatly reducing the amount of data that needs to be copied, not only saving network bandwidth resources, but also Improved the efficiency of the entire disaster recovery system;
Data recovery at any point in time: By monitoring changes in the protected data, each changed part of the data is saved in the CDP data protection zone. When the production end loses data due to failure, the data can be restored to any point in time;
Seamless switching of applications: By monitoring the status of resources such as applications and servers in real time, when an application suddenly stops abnormally or meets the conditions that require switching, the application can be automatically or manually switched to the disaster recovery server;
Platform support: Supports current mainstream Windows, Linux, and virtualization platforms, has nothing to do with servers, and supports heterogeneous servers and storage on the production side and disaster recovery side;
Fully graphical monitoring and management: All operations can be completed through the graphical interface without command line operations.
Reposted from:Baidu Security Verification