20 Sites, 80 TB: TrueNAS or ONTAP Select for Proxmox? Need real‑world input
Posted by Ready-Efficiency3090@reddit | sysadmin | View on Reddit | 11 comments
Hey folks,
I need some hive‑mind wisdom for a storage refresh across multiple sites.
We have 20 locations, each with 1–2 Proxmox hosts.
Per site we need to provide roughly 4 TB of productive SMB/NFS data, so in total around 80 TB.
The dilemma:
We’re torn between TrueNAS SCALE and NetApp ONTAP Select, but both options come with concerns.
- TrueNAS SCALE (running on Proxmox)
Concept: HBA passthrough, ZFS, backups via Veeam NAS Backup.
Concern: It runs rock‑solid on Proxmox (same Debian/KVM family), but with 20 sites I’m worried about management overhead. How realistic is it to centrally “patch things up” when something breaks?
- NetApp ONTAP Select (running on Proxmox)
Concept: SnapMirror for site‑to‑site or central backup (no Veeam needed), centralized management via BlueXP.
Concern: NetApp does not officially support Proxmox. Select is certified for ESXi and KVM on RHEL/CentOS, but not for PVE/Debian. Also, the capacity licensing for \~80 TB is a serious investment compared to TrueNAS with its flat‑rate support model.
---
My questions to you:
-
Is anyone running ONTAP Select on Proxmox in production?
-
What would you choose and why?
sybreeder1@reddit
Honestly I wouldn't use truenas as a vm in production environment. For homelab sure I used it and worked great.
malikto44@reddit
I lean to ZFS, because it has earned its bones, and in my years of using it, I've not seen data loss I could pin on the filesystem... and that is a big thing. The worst thing that happens is that ZFS hangs, and I'd rather have that than corrupted data.
benuntu@reddit
TrueNAS would be my choice not only for the cost advantage but ease of use IMHO. Very easy to integrate with Veeam and Backblaze B2. ZFS replication to another host is also a nice way to go. Not sure what you're referring to with "patch things up", but aside from updates there isn't much needed if you keep it simple. Just document your users, datasets, and permissions and train some techs on how to use it.
Ready-Efficiency3090@reddit (OP)
In this context, "patching things up" means emergency repairs or manual troubleshooting with/without a central mgmt instance
benuntu@reddit
This would be done through the Proxmox console, say if you had a corrupted boot volume, botched update or otherwise couldn't access the system. You could also SSH in if you have that enabled.
Strassi007@reddit
I would always choose Netapp. Their products are just rock solid and well designed. To be fair, i never worked with truenas.
eu_licensing_pro@reddit
Both options can work, it mostly depends on how much you value central management vs flexibility. TrueNAS is great and flexible, but with 20 sites the management overhead can become a real issue over time. NetApp is more expensive, but you get cleaner central control and less stress when something breaks. Also one thing I’ve seen in setups like this is that people focus only on storage and forget about licensing later on, especially if Windows workloads or remote access are involved. Everything can look fine at the start, but later become a problem if CALs or usage don’t match the actual setup. Just something worth thinking about early before scaling further.
Ready-Efficiency3090@reddit (OP)
Can’t you just use TrueCommand for central management and handle the site configurations with Ansible?
eu_licensing_pro@reddit
Yeah, that would definitely help with central management.
I think the bigger issue I’ve seen in setups like this is not so much the tools, but how things evolve over time across multiple sites.
At the beginning everything is aligned, but after a while access patterns change, people use systems differently, and that’s where things can drift without anyone really noticing.
That’s usually when problems show up later, not at the start.
rejectionhotlin3@reddit
ZFS is worth it.
WDWKamala@reddit
Wouldn’t the netapp solution be hundreds of thousand of dollars more expensive?
Buys a lot of labor overhead.