Running a NAS in a degraded state?

[LemonBag]

Hello all,

I was thinking about my trueNAS and I wasn't sure if I'm right or not. I figured to come to the forums to discuss more in detail.

Setup: I mainly use a 3 way mirror of 10tb drives to read data from, less commonly write to it. I backup this data to an online backup of a 2 way mirror to protect against user error deletion and whatnot.

(Boot drives: 2 ssds)

Question: If a drive from the 3 way mirror or 2 way mirror were to stop working----can I keep reading and writing from the TrueNAS as normal until the new drive gets here from RMA? Specifically, the degraded 3 way mirror zpool. I would think yes--especially since it's a mirror and no parity calculations need to be kept track of or anything. The only thing that would happen when a new drive gets here is the rebuild time would take that much longer because of the new data written to the other drives.

If not, I should still be able to use the 2 way mirror, correct?

Is there some dum-dum move I'm missing here?

 

[Arwen]

Yes, a single disk in a 2 or 3 way Mirror contains all the data and would still allow full access, both read and write. Then, when the replacement disk is available, the pool can restore it's redundancy.


Now in regards to your user error deletion comment. ZFS snapshots are one way to recover from simple user error deletions. ZFS snapshots initially take up very little space until parent dataset has lots of changes that impact the snapshot. Think of ZFS snapshots as freezing the parent dataset at the time of the snapshot. Changes, both deletes and writes can be made to the parent dataset as normal, but the snapshot remains unchanged.

Also, ZFS snapshots are a great first step at encrypting ransomware remediation. The ransomware can encrypt all the Samba shared files it wants, but as long as you have a ZFS snapshot from before, your data is still good.

ZFS snapshots almost needs its own user guide.