vdev/pool math with combined raidzX vdevs...

[Daniel Kalchev]

On 06.07.12 04:16, Jason Usher wrote:
> It's easy to find the failure math for raidz2 and raidz3.
>
> But what if you create a pool with 3 different raidz3 vdevs inside of it ?
>
> For instance, 3 12-drive raidz3 vdevs in one big pool.
>
> For each individual vdev the failure probability is now higher, since not only will it fail when 4 drives in the vdev fail, but it will also fail if four drives in any of the other two vdevs fail.
>
> So each raidz3 vdev now has a failure rate higher than vanilla raidz3 ... but what is that new failure rate ?  Is it still higher than vanilla raidz2 ?

Let's also not forget, that if a drive fails in any of the 3 raidz3 
stripes, it can be replaced independently and the sucessful replacement 
depends only on the drives in the same stripe, not all drives in the 
zpool. It also means that such a resilvering will only stress small part 
of the pool, while at the same time other drives failing in other vdevs 
will not impact your risk analysis. Further, if you have 3 vdevs, you 
could be replacing three drives at the same time, while with one single 
vdev you will only be able to replace one failed drive at a time.

Others have provided some math already.

Daniel