There are currently three common backup technologies:
- Scheduled Data Protection: Use backup software to regularly backup data. By using methods such as full backup, incremental backup, differential backup, and permanent incremental backup, data from the data center is backed up to external media for storage. At necessary moments, data from a specified time point can be restored to the production center.
- Continuous Data Protection (CDP): It can capture or track changes in data and store it separately from production data to ensure that data can be restored to any past time point. Continuous data protection systems can be implemented based on blocks, files, or applications, and can provide sufficient recovery granularity for recovery objects, achieving nearly unlimited recovery time points.
- Copy Data Management (CDM): Obtain data with application-consistent guarantees from the production environment using snapshot technology and generate a "golden copy" on non-production storage. This "golden copy" data format is the original disk format and can be virtualized into multiple copies directly mounted to servers, used for backup and recovery, disaster recovery, or development testing, etc.
The differences between the above three backup technologies can be referred to in the following table:
From the above, it can be seen that CDM backup technology has relative advantages and can be used in:
- When developing a new system, it is necessary to copy production data to the development environment for development purposes. When testing a system, it is necessary to copy production data to the testing environment for debugging. For data statistics and analysis, it is necessary to copy production data to the statistical analysis system for processing.
- Separating online data without interrupting business access, transferring queries, testing, analysis, and other traffic to non-production systems. The obtained data copy is immediately available, and the separated data is quickly provided as data support for the business through data reorganization, application pattern matching, and data cleansing.
- CDM can achieve process-oriented management and simplify the complexity of managing copy data through automated process management. CDM can also improve data management efficiency, enhance data real-time processing capabilities by combining log processing technology, and quickly handle larger and more diverse data.