ZFS deduplication is only removing some duplicates not all.
i decided to use it on some datastores such as games and so far testing i am seeing strange results.
i have copied 163GB of installed game files onto the ZFS drives games datastore from a VM in two folders under the same data store. it has saved 72GB but there is 163GB of exact duplicate files. so ZFS is removing about 44% of the duplicate files, which is nice but isn't it supposed to deduplicate all the duplicate files? or is there something else i should be doing here to make it more so aggressive?
commands used to create the games datastore on the twelve pool, poolis using blake3 checksum
also seeing similar results for backups and old vms, same OS's etc not removing 100% of duplicates just some
i decided to use it on some datastores such as games and so far testing i am seeing strange results.
i have copied 163GB of installed game files onto the ZFS drives games datastore from a VM in two folders under the same data store. it has saved 72GB but there is 163GB of exact duplicate files. so ZFS is removing about 44% of the duplicate files, which is nice but isn't it supposed to deduplicate all the duplicate files? or is there something else i should be doing here to make it more so aggressive?
commands used to create the games datastore on the twelve pool, poolis using blake3 checksum
Code:
zfs create twelve/games
zfs set dedup=on twelve/games
zfs set recordsize=256k twelve/games
zfs set compression=lz4 twelve/gamesalso seeing similar results for backups and old vms, same OS's etc not removing 100% of duplicates just some