This might sound like a dumb question, but would it be possible to disable deduplication and still be able to use incremental backups if at all?
I found this post talking about CPU usage and limiting it, there is currently no feature to do so and cpulimit is suggested.
If storage is not an issue, can deduplication be disabled to potentially save on CPU usage?
Block level deduplication and incremental backup are two features delivered by one same thing in the design of duplicacy, this is also true for tools like: kopia, borg, arqbackup, etc. So it doesn’t really make sense to ask for one without the other.
Perhaps you’re asking for file level incremental backups, this is possible for sync tools like nextcloud, mountainduck (S3), etc with their versioning feature. Or perhaps version control tools like git for text files.
There’s basically zero CPU penalty in Dupliacy’s de-dupe design, so there’d be no point disabling it even it were possible to (it’s not). Even hash lookup in memory is pretty much negligible, linear, as chunk processing dwarfs these factors.