Hey everyone and thank you in advance for any suggestions. I am newish to TrueNAS Scale. I have built a couple of small TrueNAS Core systems, but they were very simple. I want to build a more complex system and I was looking for some more insight. I have done a good bit of reading on the different vdevs and where they are applicable. I am always open to doing more if I am misunderstanding something. I am looking for feedback on what is a good/bad idea in this setup.
The setup is for a file server for a small business. There is some database usage, but majorly general file storage. When I upgrade the motherboard, I want to increase the ram and add 2 more 12TB drives to the backup vdev.
The backup vdev wouldn’t need a ton of throughput since it will be working during off hours.
But here we go.
I have:
1 x S1200btl
32 GB RAM (It’s maxed, I will be upgrading the board/RAM next year with a new budget hopefully)
8 x 12TB HDD’s (mix of 7200 RPM hdd’s)
2 x 3.84 TB SSD’s (v-nand ssd 883 DCT)
1 x 1TB nvme ssd ( MZ-VL21T0A Samsung PM9A1 1TB M.2 2280 Nvme Pcie GEN3 X4 SSD)
10GB ethernet port
I was thinking about purchasing:
1 x 320 GB nvme ssd
1 x 128 GB nvme ssd
2 x 500 GB nvme ssd’s
I was thinking I could setup:
File Server vdev:
1 320 GB ssd l2arc/slog
2 x 1 3.84TB SSD metadata
2 x 2 12TB HDD’s mirrored/striped data
Backup vdev:
2 x 2 12TB HDD’s mirrored/striped data
@Davvo Thank you for this. I wondered about the L2ARC because I always see suggestions that it’s only needed when the RAM is maxed. I have seen some mention the 64GB floor as well. Well, I am maxed and only at 32GB. So, I guess that means I don’t worry about the L2ARC until I upgrade the mobo, right?
@Stux Thank you for the insight. You are right, I was using vdev when I should have been using pool.
I have read some conflicting ideas about backups. I do want to be prepared for disk failure, but I do want to maximize storage. I guess the read/write speed doesn’t matter so much though. So, I could use the 4 disks in a Raid 5 setup.
A bit of advice: If this is for a business use, then NEVER EVER use the same machine for backup or for any type of redundancy.
Hardware can and will have a catastrophic failure at the least opportune moment and can take drives with it. I had a server do just that. A backplane hardware crash and burn took out all 4 drives in a backplane row killing one vdev in one pool and in turn the pool. Because it was a power issue it also took the server down. Now can’t power on to access the other pool. Fortunately this was easily recoverable by making some minor path re-configurations to use the backup server as the primary server while we bought and configured a new server and transferred files. No lost data and about 4 hours of lost time.