File move on ZFS truenas

[wboro]

Is there an existing request for this?

• I have searched the existing issues

Is your feature request related to a problem?

yes, moving files with mc is slow it is executing copy-delete instead of rename

Describe the solution you'd like

mc correctly detects that both source and destination are in the same ZFS pool and executes mv instead of copy-delete

Describe alternatives you've considered

No response

Additional context

No response

[ossilator]

i think mc just attempts a move, and does the copy+del dance only if that fails. so either a special variation of the rename() call is required, or this is just a manifestation of #20.

[zaltysz]

For ZFS, source and destination being on the same pool is not sufficient condition for "instantaneous" move when using typical methods. For them to work quickly, source and destination must be on the same dataset. That is because ZFS datasets are treated like separate file systems. So mc works as expected in this regard: fast move within dataset, copy-like when crossing datasets.

There is a trick to speed cross dataset copy and move on ZFS – cloning source to destination and (in case of move) deleting the source afterwards. copy_file_range() can be used for this. Although it does not guarantee cloning per se, it will do it when conditions are right: ZFS pool has block cloning enabled and if encrypted, source and destination belongs to the same encryption root.