Explanation needed for incremental backups and deduplication backups

[dpearceFL]

Since all of my backup files are the same size, can someone explain how this works or is implemented? Does it need to be enabled somewhere?

 

[Chris]

Hi,
PBS works by uploading only changed blocks as fixed size chunks (in case of VMs) or changed fragments of data as variable sized chunks (in case of LXCs and host backups). Therefore, de-duplication is performed by re-referencing already existing chunks of the datastore in the backups index file.

Technical details are described much further in the docs, see https://pbs.proxmox.com/docs/technical-overview.html

 

[Dunuin]

Since all of my backup files are the same size, can someone explain how this works or is implemented?

Where do you see this? PBS doesn`t know how big your backups are, so all sizes shown in the PBS webUI are just the size of the raw data before deduplication/compression and not how much capacity those backup snapshots actually consume of your datastore. If you want to roughly know how much space a backup snapshot consumes, have a look at the backup logs.

 

[Einar Stenberg]

There is also the "Deduplication Factor" on the summary page for each datastore that gives you a hint as to how much space is saved by dedup.
Note: You have to run a garbage collection on the store for this to update.