3 disk in pool

[SVdP Academy]

I just built a TrueNAS server using the OS drive and Storage drives. Not SCSI. just plan mounted hard drives.

It has three 1TB SATA Hard Drives.
Two are EXT4 and one is NTFS

Can I add all three hard drives into a pool? Or do I need to make three different pools?

I'm not wanting to have or build a scsi system

 

[SVdP Academy]

Second question, Can the pool be created without erasing the drives data?

 

[sretalla]

Can I add all three hard drives into a pool?

Yes.

do I need to make three different pools?

No, but you can if that's what you want.

Can the pool be created without erasing the drives data?

No. Adding disks into a pool gives those disks entirely to ZFS and they will be formatted accordingly, no original content will remain.

three 1TB SATA Hard Drives.
Two are EXT4 and one is NTFS

Although TrueNAS will work with those filesystems to perform an "import disk" operation (which copies all data off the selected disk into an already existing ZFS pool), those are not supported filesystems to use with TrueNAS functions such as sharing.

 

[SVdP Academy]

If I understand... I can have a new blank HD and make a pool with it.

Say a 4TB HHD.
Then I can import the each of the 1TB into the pool and it will copy the data from the 1TB to the 4TB.

If that is so, can I then remove the first 1TB and do the same with the second 1TB and then remove the second 1TB and live with the 4TB in the the pool?

Thanks again

 

[winnielinnie]

Drives combine into VDEVS.

VDEVS make (and expand) Pools.

Pools can then contain an indefinite number of logically distinct Datasets and ZVOLS. (Think of a "Dataset" as a "Filesystem", and a "ZVOL" as a "virtual drive". You'll likely never use ZVOLS for your use-case.)

It's highly, highly discouraged for any VDEV to be comprised of a single drive.

If you want to carry over your existing data that currently live on traditional filesystems on three separate drives (EXT4, EXT4, NTFS), then you can create a new pool large enough to comfortably house all of this data (with extra breathing room, of course), and then create three datasets right out of the gate, and then copy all of your existing data from each drive into each dataset.


Your hardware to create the pool might look something like so:

• myPool (4TB capacity)
  • Mirror VDEV
    • 4TB CMR HDD[1]
    • 4TB CMR HDD

Your ZFS tree might look something like so, considering the above 4 TB Pool:

• myPool (4TB capacity, currently comprised of a Mirror VDEV of 2 x 4TB HDDs)
  • myPool <--- top-level root dataset name, which happens to also be the same name as the pool
    • dataset1 <--- copy everything from the 1TB EXT4 drive in here
    • dataset2 <--- copy everything from the other 1TB EXT4 drive in here
    • dataset3 <--- copy everything from the 1TB NTFS drive in here

Or if you want to use the "pseudo-root" method, which I use to spare myself from future headaches, and for more intuitive management:

• myPool (4TB capacity, currently comprised of a Mirror VDEV of 2 x 4TB HDDs)
  • myPool <--- top-level root dataset name, which happens to also be the same name as the pool
    • myData <--- pseudo-root that is a parent to your important data
      • dataset1 <--- copy everything from the 1TB EXT4 drive in here
      • dataset2 <--- copy everything from the other 1TB EXT4 drive in here
      • dataset3 <--- copy everything from the 1TB NTFS drive in here
    • tempData <--- pseudo-root that is a parent for temporary or unimportant things
      • datasetPlayground
      • datasetBigTempDownloads

Once you've copied everything over, you can then decide what you want to do with your original three 1TB drives. Leave them as they are? Use them for backups? Combine them into a VDEV to expand your existing pool? to create a new pool? to use as a new pool in another system for backups? Etc, etc, etc. (Backups are super crucial, so always make sure you have multiple copies of important data in more than one location.)

[1] Avoid SMR drives (use only CMR) when deciding on a purchase.