Hi, I recently put together a TrueNAS Scale system with my old X370 motherboard, a new AMD Ryzen 5 5500, 16gb of DDR4 ram capable of 3200 MHz (left this stock for now, unsure if I should change anything here) and a 256gb NVME boot drive. Currently, I have a mirror with 2x24TB Seagate Ironwolf Pro drives. Should I create a new pool when I add my other (older) pair of Seagate Skyhawk Al 24TB drives to the mix, and simply create another 2x24TB mirror? Or would it be better to add them on to my existing 24TB pool? I’m very new at this and have only used Windows Storage Spaces previously, sorry for the noob questions
Either could be fine. If you add them to your main pool as a second mirror VDEV, you will have a single pool with all that space and the performance of the two VDEV configuration.
If you make a new pool with the two drive mirror configuration that may help if you wanted to keep them segregated more. Might be advantage. If one pool dies, you don’t lose all data, just the single pools
MAIN_POOL (gives you a single pool with aprox 48TB contiguous space and faster read and writes)
VDEV1 2x 24TB Seagate Ironwolf Pro
VDEV2 2x 24TB Seagate Skyhawk
Or this
MAIN_POOL
VDEV1 2x 24TB Seagate Ironwolf Pro
SUR_POOL
VDEV2 2x 24TB Seagate Skyhawk
BASICS
TrueNAS Systems pool layout whitepaper
White Papers | TrueNAS - Open Enterprise Storage look for ZFS Storage Pool Layout
3 Likes
See each vdev as a virtual disk. Losing any of the vdevs will result in pool failure. Increasing the number of disks in a mirror will improve reliability, increasing number of vdevs will increase capacity.
Mirrors = single disk capacity with duplications, limited storage
Strip = data distributed across all disks, max storage, 0 fail safes
Raidz1 = capacity -1 disk, 1 disk safety
Raidz2 = capacity -2 disk, 2 disk safety
Raidz3 = you get the idea
You can have a vdev of the above configuration and a new vdev matching redundancy level there after. Losing any of the vdevs will result in pool failure so configure as you see fit.
1 Like
Now i will answer with a question…
What is your use case? What do you except to do? How important is your data?
What flaged my attention was used hard drives. How used?
I hope you are doing better since the broken bones.
1 Like
They were bought new by me in may 2024, used as a windows storage spaces mirror since then
Have you performed any burn-in tests on those drives? I’m asking because if you start to use them, they could start failing with ZFS having to rewrite data or just start throwing errors.
At a minimum run a SMART long test on each drive. 24tb could take a week to complete. That means uninterrupted power until complete.
If your data is important then redundancy becomes very imporant. That is why i asked that question.