Jbod back plane connector issues.

[Gagik]

Hello,

I'd like to discuss a scenario with you and seek your insights into the potential consequences if it were to occur.

I have a 60-bay unit equipped with 60 drives. The setup involves 5 vdevs configured as RAID Z2. The drives are organized in the following order:

Slots:
0, 1, 2, 3, 4 Vdev1 Raidz2
5, 6, 7, 8, 9 Vdev2 Raidz2
10, 11, 12, 13 Vdev3 Raidz2
And so on...

Now, let's consider a situation where the backplane of the unit experiences issues, causing drives 0, 1, and 2 to go offline. Additionally, let's assume that drives 10, 11, and 13 also go offline due to the same issue. It's important to note that the drives themselves aren't failing; rather, it's the backplane that is faulty. My question is: How would this impact the volume? If I were to obtain a new unit and transfer the drives to it, would I be able to restore the volume to a functional state? Or would this lead to a significant failure of the volume or other complications that need to be addressed?

 

[sretalla]

0, 1, 2, 3, 4 Vdev1 Raidz2

drives 0, 1, and 2 to go offline.

Your pool is degraded and offline due to insufficient replicas of that first VDEV... only further compounded by the other drives that may be offline.

If I were to obtain a new unit and transfer the drives to it, would I be able to restore the volume to a functional state?

Yes, all good.

Or would this lead to a significant failure of the volume or other complications that need to be addressed?

No, ZFS protects for exactly that by taking the pool offline.

 

[Gagik]

Thank you for the feedback. I can rest now lol..