On Tue, 10 Mar 2009, A Darren Dunham wrote:
What part isn't true? ZFS has a independent checksum for the data
block. But if the data block is spread over multiple disks, then each
of the disks have to be read to verify the checksum.
I interpreted what you said to imply that RAID6 type algorithms were
being used to validate the data, rather than to correct wrong data. I
agree that it is necessary to read a full ZFS block in order to use
the ZFS block checksum. I also agree that a raidz2 vdev has IOPS
behavior which is similar to a single disk.
From what I understand, a raidz2 with a very large number of disks
won't use all of the disks to store one ZFS block. There is a maximum
number of disks in a stripe which can be supported by the ZFS block
size.
--
Bob Friesenhahn
bfrie...@simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/
GraphicsMagick Maintainer, http://www.GraphicsMagick.org/
_______________________________________________
zfs-discuss mailing list
zfs-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
