vdev/pool math with combined raidzX vdevs...
Jason Usher
jusher71 at yahoo.com
Wed Jul 11 15:32:41 UTC 2012
Hi Bob,
--- On Wed, 7/11/12, Bob Friesenhahn <bfriesen at simple.dallas.tx.us> wrote:
> The probabilty is indeed additive just as you say. My
> point is that the fundamental integrity is offered at the
> vdev level. If a vdev fails, then the whole pool is
> gone. The MTTDL calculations for various vdev
> topologies vary by orders of magnitude, which tends to make
> the additive nature of more vdevs insignificant.
Thanks again for responding.
I'm not going to beat this to death, but just to summarize, if F is 2, then the corresponding data loss probabilities for RAID-Z1, -Z2, -Z3 are: 14.9%, 1.3%, and 0.086%.
But if combining multiple vdevs into a zpool (as opposed to maintaining a different zpool for each raidz3 vdev) is additive, then raidz3 becomes .258%.
Since (I think) a lot of raidz3 adoption is due to folks desiring "some overkill" as they attempt to overcome the "disks got really big but didn't get any faster (for rebuilds)"[1] ... but they are losing some of that by combining vdevs in a single pool.
Not losing so much that they're back down to the failure rate of a single raidz*2* vdev, but they're not at the overkill level they thought they were at either.
I think that's important, or at least worth noting...
[1] http://storagegaga.com/4tb-disks-the-end-of-raid/
More information about the freebsd-fs
mailing list