NanoBSD on ZFS

I have a request to make a NanoBSD image for a user using a ZFS file system. Has anyone ever used ZFS on a NanoBSD build?

Personally it seems like overkill. NanoBSD runs UFS as Read-Only so I really don't see the utility from a robustness factor.

Am I thinking wrong?

What extra tools will I need to build in? Do base tools cover ordinary usage OK?
 
Revisiting this topic I am looking at ZFS docs and all I see is complexity on top of complexity.
For a single disk what is the advantage of ZFS over UFS?
From my reading the main benefit is snapshots?
 
For a single disk what is the advantage of ZFS over UFS?
- Bulletproof upgrades with ZFS Boot Environments (sysutils/beadm).
- Data integrity using checksums for blocks.
- Fast Compression (lz4) for 'free'.
- Flexibility in data management (once You do UFS partitioning You will not resize and will not increase inodes).
 
As a side note UFS supports snapshots, too (albeit with some limitations; see mount(8), search for snapshot).

However, for a NanoBSD system, where root file system is (mostly) immutable and, if you prepare your image right and create a secondary root file system partition, upgrades are easily atomic, as well.

You might consider ZFS for a data partition on a NanoBSD system, where you store run-time data and need persistence (database files, for example). Regarding extra tools, that depends on your use case: if you use a 3rd party packages just include them in your NanoBSD system.
 
Back
Top