Currently, backups for S3 are done in full each time and this requires large data storage on S3 which translates into costs. Despite everything that S3 is cheap, I think it can be optimized.
S3 is a very good solution for backups, which can be in addition to btrfs block volume backups.
It primarily increases the resilience to failures and data loss.
Is it also possible to implement incremental backups to S3?
I remember that Plesk used this so it is rather possible.