Restrict cloning with different properties

While technically its not a problem to clone between datasets with
different properties, it might create expectation of new properties
being applied during data move, while actually it won't happen.
For copies and checksum it may mean incorrect safety expectations.
For dedup, compression and special_small_blocks -- performance and
space usage. New zfs_bclone_strict_properties tunable controls it.

Reviewed-by: Rob Norris <robn@despairlabs.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Alexander Motin <alexander.motin@TrueNAS.com>
Closes #18180
This commit is contained in:
Alexander Motin
2026-02-10 12:53:24 -05:00
committed by GitHub
parent 1412bdc6c2
commit aa29455dd7
17 changed files with 193 additions and 24 deletions
+1
View File
@@ -110,6 +110,7 @@ VOL_RECURSIVE vol.recursive UNSUPPORTED
VOL_REQUEST_SYNC vol.request_sync zvol_request_sync
VOL_USE_BLK_MQ UNSUPPORTED zvol_use_blk_mq
BCLONE_ENABLED bclone_enabled zfs_bclone_enabled
BCLONE_STRICT_PROPERTIES bclone_strict_properties zfs_bclone_strict_properties
BCLONE_WAIT_DIRTY bclone_wait_dirty zfs_bclone_wait_dirty
DIO_ENABLED dio_enabled zfs_dio_enabled
DIO_STRICT dio_strict zfs_dio_strict