Hello everyone,

I recently updated my Unraid server to the latest version 6.12.0 and am now in the process of reconfiguring my existing array. So far, I've been using the Dynamix File Integrity plugin to monitor my files and noticed that every few months there are reports of corrupt data, interestingly, mostly with RAR archives.

 

Because of this, I'm interested in ZFS as it offers the ability to "repair" data automatically, without me having to manually retrieve it from the backup. I have some questions about this:

 

• How do individual HDDs in the array with ZFS compare to XFS in terms of data integrity and repair?
• Would it make sense to rely more on ZFS pools, possibly even smaller ones, like three pools each with 4 HDDs in a RAIDZ1 configuration, to save energy? What are the pros and cons I need to consider here?
• Is it possible to integrate various ZFS pools and individual HDDs into the same array? I want to have the option to push data equally onto my shared folders without a pool filling up and thus preventing the respective share from taking in more data. If this is not possible, how do I proceed when a ZFS pool is full?
• Can I move data between the ZFS pools, and if so, how would this work?

 

I look forward to your advice and suggestions for optimizing my setup. Thank you in advance for your help!

Thanks,
Cout99

1 hour ago, Cout99 said:

How do individual HDDs in the array with ZFS compare to XFS in terms of data integrity and repair?

They can detect but not repair data corruption.

 

1 hour ago, Cout99 said:

Is it possible to integrate various ZFS pools and individual HDDs into the same array?

You cannot have multi device zfs pools in the array, see here for some advantages and disadvantages of using zfs in the arrays vs pool.

1 hour ago, Cout99 said:

Because of this, I'm interested in ZFS as it offers the ability to "repair" data automatically, without me having to manually retrieve it from the backup. I have some questions about this:

The repair capability only applies when the drives are in a redundant ZFS array.   When used in the array then each ZFS drive is a single drive system that can detect but not repair corruption.   In this respect is similar to BTRFS drives in the main array.