Continuous Data Protection

Continuous data protection is a backup system that allows administrators to keep a near identical version of a database ready for deployment. It also provides disaster recovery as a feature the process. There are many different types of processes for performing a backup and many companies that provide software products to do so.

Backup systems primarily use one of two common levels. One is file-level storage, the other is block level. Both have advantages and disadvantages, mostly related to the speed of transferring data.

File level storage tracks data by folders and files. A backup system using file level processes tracks changes to individual files and folders. The backup software has to maintain the addresses and b-trees for each file and folder. This process tends to be slower, but easier to maintain.

Block level storage tracks data by blocks of memory. These blocks can be treated as separate hard drives. Instead of changing data by folder and file, if there is a change in any data in a block, the whole block is backed-up. This process is faster, but requires more maintenance and control.

There are three primary systems for backing-up data, as well: full, incremental, and differential.

• A full backup is when all files are copied to the backup site in one transaction regardless if they have changed or not.
• A differential backup updates only those files and folders that have changed since a full backup was performed.
• An incremental backup only changes those files and folders that changed since the last incremental backup was performed. The initial backup will can be a full one or of only certain files and folders, but thereafter, the only backups will be those files that changed, regardless of how long ago a full backup was performed.

Software companies that produce backup products will use different blends of levels and processes. Their products may also provide options that allow an administrator to choose the combinations that are most appropriate for their network environment.

Continuous data protection is more than just backing-up data, however. Continuous data protection uses back-up methods but creates redundant copies of a database to be restored in case of corruption and for fast disaster recovery. It typically involves several layers of servers and tape systems, provides an instant mirror of the database, and various options for restoring a database.