Pools Datasets Snapshots - use and recommendations

Hello i am new to Truenas, i use it for homelab.

Juste created 1 pool + 1 dataset (raidz1, 6 wide, 16tb*6) and threw 20tb of files there.
Then organized everything just using folders and subfolders (inside my main and only dataset), then shared some folders via SMB.

I am new to ZFS, never usez pools, datasets and snapshots.

I have :
-media files for Plex (many small music files + many larger video files)
-shares for users from linux/pc/mac (they have one folder per user and throw anything there to backup files mainly)
-windows backups (via integrated backup option in windows)
-proxmox VMs backups (PBS)

When / why to crate new pools ?
When / why to create new datasets ?
Why/what/when do you snapshot ?

Does it make sense to snapshot everything and only having one dataset ?
Is there recommended practices to use datasets and snapshots ?

Best regards :slight_smile:

No comments for your datasets and snapshots.

Raid-Z1 on 6x 16TB drives isn’t recommended due to resilver time with large drives

Raid-Z2 would be safer, but it will lose capacity on your pool

1 Like

Different hardware needs: Bulk storage (raidz#) vs. block storage (mirrors).

Different protocols (SMB, NFS, iSCSI). Different users. Different uses, and different settings for different content (record size, compression,…).

For safety (user errors, ransomware). As often as you’d need if you ever had to restore, and retained as long as is useful (and then some).

1 Like

Snapshots:

1 Like