Nathan Essex wrote:
Thank You, so that means that even if I use something that writes raw i/o to a
zfs emulated volume, I still get the checksum protection, and hence data
corruption protection.
yes it does.
Also consider how BAD performance could be if it were actually
calculated on a per file basis.
For example a 1 bit write on to the end of a 5G file would require
reading and checksuming 5G of data to calculate the new checksum.
Block level is the only sensible way to do this IMO.
--
Darren J Moffat
_______________________________________________
zfs-discuss mailing list
zfs-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/zfs-discuss