Hey all!
I wanted to check in to see what members of the forum will be using their HL15 for, are you getting a full build? chassis + backplane/PSU?
What will you being using your HL15 for?
I plan on migrating the components I have in my Proxmox homelab to a chassis+backplane model
Excited to hear everyone’s plans and maybe utilize some in my own environment!
This will be the push to finally buy a rack too. After watching the video from @technotim , I was sold. I had been looking at all kinds of cases and hardware to build my own system but I was convinced that buying a system built by experts was the way to go.
I plan to add it as another node to my ProxMox cluster with TrueNAS in a VM and, one day, add a second HL15 
Part of the beauty with the HL15 is the ability to use it as a tower or rackmount, so you’ll have plenty of time to come up with a rack!
I want the excuse 
It’ll be added to this project as a fourth node. 
My HL15 will be a storage array for my UptimeLabs Raspberry Pi cluster. Currently taking bets on which will arrive first 
I have a server that serves both my compute and storage needs. I have 31.5TB (48TB Raw) of usable storage. I plan to replace and resilver almost all the drives that I have right now as I am already filling that up. A few TB is used for my personal deep storage needs and archival purposes, however the rest is consumed by my Plex server. I run a 300+ user plex community and run a server that I normally have all slots full on.
I plan to use the HL15 to remove the storage element from my compute and have TrueNAS running on it’s own machine. Right now I virtualize it within Proxmox.
Will be performing a chassis swap for my main unRAID server 
Im doing a chassis swap for my unraid server… I really want to switch over to proxmox for VM usage, but cant seem to figure out the best way to do that while also having a massive software raid share… I currently use my Unraid for Docker, VMs, Storage and Jellyfin. what would you guys think to be the best way to go about this, im about at the point of just nuking unraid and doing proxmox with a TrueNas guest.
I have an Nvidia GPU that i was able to get working with vGPU support in proxmox and I really want to take advantage of that.
Hi, @Lavavex Proxmox is a great choice for an OS, It even has native ZFS built in.
I would recommend making a ZFS pool for your storage and then you can simply make a share with that storage (SMB/NFS). You can even install our Houston UI on Proxmox 7 to help you manage the shares and the ZFS pool
This way you also have the Proxmox OS for all of your VMs and computing needs while having a bare metal storage RAID
Combined home server + NAS. Replacing two separate devices right now. Will run TrueNAS and maybe one VM.
To replace my ageing but trustworthy TrueNAS server which these days is primarily a backup target, but also runs a couple of VMs. I haven’t yet decided if I will run TrueNAS baremetal or virtualized on the HL15, or if I will try Houston UI instead of TrueNAS.
So, install houston as a vm, or as a proxmox plugin or something? sorry, im kinda confused as I have never touched proxmox or houston before (im a unraid script kiddie)
Also I have mismatched disks due to using unraid. I have 2 4tb and 1 6tb drive with an 8tb parity drive… how screwed would I be if i want to switch over to proxmox/houston?
Hey @Lavavex, with ZFS the drives are grouped into VDEVs which are then stripped together to make the ZFS pool
Each drive included in a VDEV should be the same and if there is any type of mismatch the VDEV will use the least common denominator.
With that, if you were to use the drives you currently have the drives would only utilize 2TB of them giving you 8TB of RAW useable space.
As for Houston is it a UI which you can add to your Linux server allowing you to manage your server from the browser. This web UI can be installed on Rhel and Debian based operating systems.
I think I see where there may be a bit of a disconnect. To add to what @Hutch-45Drives mentioned regarding VDEVs in ZFS, this is going to be a bit of a change for you coming from unRAID. The most common file system used in unRAID is BTRFS, which plays a large role in unRAID’s appeal to many Self-Hosters!
BTRFS is a file system that allows you to create a parity drive (using your largest drive available to the system) to sort of “checksum” or tally up the entire contents of the drives below. To the best of my ability to explain that, and probably grossly oversimplifying the features of BTRFS for sake of easier explanations.
We know that data is stored as 0 or 1, in bits. We also know that you have 3 capacity drives in your array (as seen in the “Main” tab) and your 8TB parity drive.
So, those drives may look something like this:
| Drive 1 | Drive 2 | Drive 3 | Parity |
|---|---|---|---|
| 0 | 1 | 0 | 1 |
| 0 | 0 | 1 | 1 |
| 1 | 0 | 0 | 1 |
| 0 | 0 | 0 | 0 |
| 0 | 1 | 0 | 1 |
| 0 | 0 | 1 | 1 |
| 1 | 0 | 0 | 1 |
| 0 | 0 | 0 | 0 |
| 1 | 1 | ||
| 0 | 0 | ||
| 1 | 1 | ||
| - | |||
| - | |||
| - |
In this diagram, Drive 1 and 2 are your 4TB’s and Drive 3 is your 6TB, which is why the string is longer! I kept the binary math simple here to just convey the point. This is all just to show that the Parity Drive is just all of your array drives “added up” and the total is written to parity. This is why when you lose any one of your drives, you can rebuild from parity.
Drive 2 dies? No problem! We know what Drive 1 and 3 contain, as they are online and healthy. So we subtract their values from the Parity drive and know what Drive 2 MUST have contained. We then simply write that answer to the new drive you replace it with! The largest benefit here is we can keep expanding and adding more drives to the math problem with ease over time. This makes BTRFS really nice for a use case such as a self-hosted media server, ease of expansion.
Now, ZFS on the other hand, does not function the same way by using a parity drive. Instead, it handles your data integrity through a myriad of different options, such as RAIDZ1/2 or a simple Mirror. The problem with this is that unlike BTRFS, it’s required that the drives match in size and that you cannot as easily add capacity to a Mirror or RAIDZ array like you can with BTRFS.
So, it may be in your use case, that BTRFS is the file system best suited to your needs! And that’s fine! There are tradeoffs with any solution.
A gentleman much smarter than I could ever hope to be has a video here: https://youtu.be/lsFDp-W1Ks0?si=KMrIFNRotUm7SwDu (If I’m not mistaken, Wendell was part of the committee helping with the 45Homelabs project)
Thanks @orix for that great explination of the differences between ZFS and BTRFS that was an excelent example!
ok… i know unraid has support for mismatched disks in zfs, so i have no clue how they are doing it then…
I will probably stick with unraid for now once i get the case just so I dont have to buy new drives right out the gate. but am fine with buying new drives for the case to switch it over to a ZFS pool with proxmox and houston. just need to figure out the best way to move all my docker containers over.
Or ill just use proxmox and have my unraid usb connected to a vm just for unraid…
I’m planning on using it for a personal Plex server, steam cache, home assistant, running truenas scale. Making it portable for the frequent work trips that feature 6 months with limited connectivity.
nice! did you know, steam now has a steam cache like function built in now, just have a VM with steam installed or a docker container with steam installed and it will do all the work for you!
Kubernetes/Ceph node
Gonna replace my 12 yr-old 12 bay synology server. I’ll be trying out Houston UI but I really need resilio and a few other maintenance packages so we’ll see how it goes.
Otherwise I’ll look into https://xpenology.org/ which seems to have a way to run Synology DSM in non-Synology systems