When deduplication is most effective
The following are cases when deduplication produces the maximum effect:
- When backing up in the full backup mode similar data from different sources. Such is the case when you back up operating systems and applications deployed from a single source over the network.
- When performing incremental backups of similar data from different sources, provided that the changes to the data are also similar. Such is the case when you deploy updates to these systems and apply the incremental backup. Again, it is recommended that you first back up one machine and then the others, all at once or one by one.
- When performing incremental backups of data that does not change itself, but changes its location. Such is the case when multiple pieces of data circulate over the network or within one system. Each time a piece of data moves, it is included in the incremental backup which becomes sizeable while it does not contain new data. Deduplication helps to solve the problem: each time an item appears in a new place, a reference to the item is saved instead of the item itself.
Deduplication and incremental backups
In case of random changes to the data, de-duplication at incremental backup will not produce much effect because:
- The deduplicated items that have not changed are not included in the incremental backup.
- The deduplicated items that have changed are not identical anymore and therefore will not be deduplicated.