Best practices - Datasets & shares for home office setup

[eysimir]

Hello community!
I recently completed the build and installation of a TrueNAS PC with one Samsung 870 EVO 250GB SSD for OS and two mirrored Seagate Exos X16 12TB disks for storage. I have read the documentation and installation guides an I have reached the point where I need to decide the hierarchical layout of my datasets and shares. I know this is up to me to decide, but I'd love to hear some opinions and comments about what you think are the best practices in general, given below circumstances.

Usage:
I plan to use the NAS primarily for home office Windows backup purposes but I would also like to have a public LAN area that can be accessed by macs, phones & tablets et.c. on my network. I might also want to access the server remotely over the internet, but this is not the highest priority.

Datasets:
I currently only have the root dataset of the mirrored pool called "exos" and I know I need SMB datasets for Windows (and Mac) access. I think I want private shares for individual users so I can see a couple of options here. Either simply create individual datasets for each user directly below the root dataset, or make a parent SMB dataset under which all individual datasets would reside. What would be the pros and cons of either one?

Code:

exos
  user1
  user2

Code:

exos
  smb
    user1
    user2

Permissions:
I find permissions a little confusing. Any suggestions here? What is the main reason to use ACLs and not just the basic Permissions Editor?

Shares:
I suppose that once I have my datasets in order, I would simply create one SMB share for each dataset. What other shares could one need? I'd love some input here!

I'd be happy for any comments you have regarding this setup.
Thanks!
/Anders

 

[NugentS]

I use your second example
ie I have a top level data set called SMB, one called NFS, one called iSCSI etc. It helps me keep thing organised in my own mind. Under SMB I have the "usual" common, install, etc

 

[eysimir]

Thanks! From a mental perspective, I clearly understand your separation of each type of dataset. It makes total sense. I don't yet know if it would make sense for my setup, but I see it works for you.

I might reveal my lack of knowledge here, but what are you using NFS and iSCSI for? Also, I have no idea of what the "usual" is. Would you care to explain briefly? It's my first time building a NAS, and although I have a computer science background, I'm not very experienced in network stuff.

Thanks a bunch!

 

[NugentS]

I use iSCSI & NFS - still trying to decide which one I prefer - to Virtual Disks for VMWare. I have a home lab.
I also have a dataset called snapshots for when I replicate from one pool to another (I have a small pool of SSD's as small fast storage).

If I was using home drives - which I keep meaning to try I would probably first try with a dataset called home or homes and any home datasets underneath that

 

[NugentS]

Oh and "Usual" = whatever fileshares work in your setup, common, video, install, music, etc