Hello All, I'm fairly new to UnRaid OS and new to ZFS file system as well. I currently have 6 4TB hard drives and planning on getting more. What i'm trying to decide is how do I get the most performance, storage space, and fault tolerance from my current drives in a ZFS pool but also make future expandability of the pool easy?

 

For the ZFS pool should I do 1 vdev in raidz2 with all 6 drives, 2 vdevs of 3 drives each in raidz1 or 3 vdevs of 2 drives each mirrored ?

 

Overall, I would like the ZFS pool to be able to recover from 2 drive failures. Hoping the great minds here can help me figure this out.

Just wanted to add another related question for anyone who can answer: In the case of a ZFS pool with multiple vdevs of drives in raidz1, can each vdev survive 1 drive failure or the entire ZFS pool can only survive 1 drive failure ?

16 minutes ago, Mephisto76 said:

Overall, I would like the ZFS pool to be able to recover from 2 drive failures.

That would require a raidz2 pool with all the disks.

 

9 minutes ago, Mephisto76 said:

In the case of a ZFS pool with multiple vdevs of drives in raidz1, can each vdev survive 1 drive failure or the entire ZFS pool can only survive 1 drive failure ?

You can loose one device from each vdev, but if you lose two devices from the same vdev you will lose the complete pool.

1 hour ago, JorgeB said:

That would require a raidz2 pool with all the disks.

 

You can loose one device from each vdev, but if you lose two devices from the same vdev you will lose the complete pool.

So if you were me (with 6x 4TB HDDs), would you configure a ZFS pool as 2 vdevs in raidz1 each ?

I would make a single raidz2 vdev.

2 hours ago, JorgeB said:

I would make a single raidz2 vdev.

With just one vdev consisting of 6x 4TB HDDs, when I want to expand the ZFS pool in the future, I would have to add 6 more drives of the same size, right ?

Correct, at least until raidz expansion is available, which is expected in about a year.