> > So the question becomes, what are the tradeoffs > > between running a two-way > > mirror vs. running RAID-Z on two disks? > > A two-way mirror would be better -- no parity > generation, and you have > the ability to attach/detach for more or less > replication. (We could > optimize the RAID-Z code for the two-disk case, but > all that would do > is make it into a two-way mirror under the covers.)
Thanks; getting real info is so much nicer than guessing! > > I had hoped to be able to add additional disks into > > an existing RAID-Z > > and expand the available space, but that doesn't > > eem to be possible > > ZFS uses dynamic striping. So rather than growing an > existing RAID-Z > group, you just add another one. That is, suppose > you create the pool > like this: > > zpool create tank raidz disk1 disk2 disk3 > > To add more space, you'd say: > > zpool add tank raidz disk4 disk5 disk6 > > The down side is that you can't add just one disk -- > you have to add them > in small groups. The upside is that by doing it this > way, there's no > need for block remapping -- so adding space isn't a > violent act. It also means that the 4- and 5-disk hot-swap chassis give you zero realizable expansion -- at least if you're using a minimum of three disks for a RAID-Z group. Starting a new group costs you the parity overhead again, too; but then expanding the group too much means surviving a single-drive failure isn't good enough, either. Which means going to the much-more-expensive 8-drive chassis. Having adding space not be a "violent act" certainly has a LOT to recommend it. Anyway, thanks very much for the authoritative answers. This message posted from opensolaris.org _______________________________________________ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
