XigmaNAS

vdevs, pools, datasets, volumes

Can someone clarify these and why I need them?

I found I only needed a vdev and pool to start using it, so what's the point of the dataset or volume? I'm new to ZFS, so use small words.

I'm not sure if it helps, but my needs are a basic file server using AFP, SMB, FTP, TFTP, RSYNC, and maybe NFS over a single gigabit connection. I have 10x 4TB drives in a single RAIDz2 with 16GB ram and no desire of SSD silliness(w/ ZIL/L2ARC), but the system does have a SSD that I could allocate some space if it would help greatly. I am definitely never adding another 10 disc vdev, as any upgrade will be a separate system or wiping and redoing this one.

Thanks!

please read this powerpoint, is really usefull: http://forums.freenas.org/threads/slide ... oobs.7775/

Thanks, that does help some.

However, they merely mention that zpools are sometimes referred to as volumes, and do not mention datasets at all. Can someone clarify these two? and why/if I need them?

There are sections in the WebGUI for making them, but I can't figure out the difference or reasons for either.

BTW, a reference to that post (or something like it more related to N4F) should be in a sticky.

(http://www.trainsignal.com/blog/zfs-nas-setup-guide)

summary

A pool or volume(this is the same) can be shared on the network, or it can be optionally divided into ZFS datasets. A dataset is like a folder on the volume, but it acts like a filesystem in that it supports snapshots, quotas and compression. Once a dataset is created its permissions can be set independently to that of the ZFS volume. This means that several datasets can be created, one for each group of users (eg. sales, developers, and marketing).

Thanks, that helps a lot.

So, is this correct?

vdev = Virtual DEVice (single disk or group of disks in stripe, mirror, raidz1, raidz2, or raidz3)
- you name it, pick the disks, and how(if) to group them

pool = single or multiple vdevs
- you name it, pick the vdev(s), and optionally change the mount point

volume = same as a pool but with ability to use compression(on the whole pool) and dedup
- you name it, pick the pool, specify the size (why would you specify anything other than the entire size?), optionally turn on compression, optionally turn on dedup
- So, the only reason to make this is to have the whole pool compressed or deduped? (rather than per dataset)
- If I specify a smaller size, can I make more than one volume on a single pool? This would be similar to the function of a dataset, but without permissions and quotas?

dataset = like a folder, supports snapshots, quotas and compression (on the dataset), and permissions
- you name it, pick the pool, optionally turn on compression, optionally turn on dedup, optionally set quota and permissions
- So you can make several datasets on a single pool, right?
- Can I oversubscribe quota amounts (e.g. make 5x datasets each with a 1tb quota on a single 4tb pool)?

There is also explanations of these terms on the wiki

http://wiki.nas4free.org/doku.php?id=zfs:main