45HomeLab HL15 vs TrueNAS Mini R

I’ve been looking at either the “Fully Built & Burned In” HL15 from 45HomeLab or the TrueNAS Mini R. The main purpose will be storing backups and media files.
Here are my priorities in no particular order:

  • 12+ 3.5in hot-swap HDD bays
  • Low(ish)* power
  • Quiet(ish)**
  • ZFS
  • ECC memory
  • 10gb networking (SFP+ preferred)
  • Rack mounted
  • Reliable/Tested
  • Non-proprietary hardware

*<175W w/ drives
**Barely audible through closed door

Looking for some feedback and/or recommendations. Thanks in advanced. :slight_smile:

1 Like

I had originally looked up both of the CPU’s benchmark scores and figured their TDPs were comparable. Looking them up on Intel Ark I can see that the HL15’s Xeon Bronze 3204 has a TDP of 85W whereas the Mini R’s Atom C3758 is 25W.

I know TDP != power consumption, but I figured I should look further into each platform’s power consumption figures. Here’s what each company publishes.

Idle (empty) Idle (full) Load (full) Startup (full)
HL15 120W 280W 350W 510W
Mini R 63W N/A 167W* N/A

Also looking at each platform’s noise levels. I don’t think dB and dBA are directly comparable though?

Idle Peak
HL15 32.5dBA 32.5dBA
Mini R 45dB 53dB

Fully loaded we saw 89W at idle in our test unit. With the WD Red Plus 10TB drives, optics, CPU, longer 10Gbase-T runs, and more under load we were able to exceed 200W in this system on 120V power.

*ServeTheHome was able to get the TrueNAS Mini X+ to over 200W. I’d be surprised if the Mini R uses less with 2x+ drives?

Struggling to find anybody reviewing either of these products and providing this type of information.

2 Likes

The aim for different markets. HL-15 is more leaning towards DIY people and TrueNAS Mini is a great alternative to prebuilt NAS systems from QNAP or Synology.

HL-15 is more freestyle, more performance, more flexibility/tinkering, more homeserver.
TrueNAS Mini is your (comparably)vcheap and carefree ZFS NAS appliance.

DO you want homelab or do you want a NAS?

The TrueNAS Mini R is $1850. The HL15, fully built, is $2000.

This would be a part of an overall bigger homelab that I originally built using 2nd-hand enterprise gear I cobbled together a few years ago. At this point I’m looking for a reliable, low power, and quiet network storage system.

Here’s the goal; I’ve got 6(+1 spare) Seagate Exos X20 20TB drives that I plan to throw into a single RAID-Z2. Performance isn’t paramount, basically all the data is large sequential files like backups and media… This covers me for the next few years, but I want to be able to increase the pool size.
I don’t intend to run anything more than MinIO on the server. The sole purpose will be hosting NFS, SMB, and S3 shares.

Typical object storage workload…I’m not sure how CPU intensive MinIO is, but I doubt any of the two products will have any problem running dual 10G connection for file share or object store.

HL15 is the better one here. 15 bays for plug&play hotplug is the winner here. And it has free PCIe slots for HBA expansion. Being a standard server chassis and board, it’s more flexible in terms of expansion. And you can upgrade CPU and memory very easily too

2 Likes

That’s my thought… but if it uses 2x the power of the Mini R I don’t think it’s worth it. I can live with something like an additional 60W for the same workloads. I found someone giving some power usage numbers for the HL15, but still looking for someone with a Mini R.

The Mini R appears to use a standard Supermicro board with an embedded Atom, but RAM is upgradeable and it has a free PCIe slot i could throw a HBA or a M.2 riser card. So it’s not completely locked in. The HL15 is definitely more flexible though.

This is going against my intention of getting a system that is fully verified, but I guess I could get the HL15 chassis/backplane/PSU and some ASRockRack/Supermicro board with an embedded Atom or Epyc to have lower power usage… My intention is to run TrueNAS Core. This would mean I’d want the board to have excellent FreeBSD support.

Just an observation that might or might not be helpful. I have a couple of the AV-15 chassis (similar to the HL15’s) and I have no problem keeping my large capacity drives cool with minimal noise with a selection of 120mm fans to choose from. While I don’t have any experience with this particular 2U chassis from ixsystems I have, in the past, struggled to control the noise level with other 2U chassis when trying to cool a large number of 3.5" HDDs when the activity level starts to rise. This is in the context of a homelab enrvironment. Just a thought.

1 Like

Generally more U leads to less fan noise because you won’t need to have the 40mm or 80mm screamers.

+1 vote for HL15 chassis + backpane only if the alternative is the 2U Mini R.

1 Like

That’s what I was thinking, but iX Systems is claiming the Mini R is fairly quiet. I’m sure both will be quieter (and more efficient) than the current NetApp disk shelf and HPE ProLiant DL380e gen8. I’ve reached out to iX Systems to see if they can provide a dBA sound rating for the Mini R, I don’t expect to hear anything until next year.

It’s not the only alternative, if there are any other suggestions that are around $2000 for everything except drives that also fit my posted priorities.

I’ve just looked through Supermicro’s and ASRockRack’s sites and not seeing any embedded CPU options that have >=15 SATA connections and are under $1000. Most are ITX and I don’t want to waste the single PCIe slot on an HBA.

Here’s the response I’ve received from iXsystems:

Thanks for your patience. The platform team got back to me with the following:

We’ve measured the Mini R (populated with drives) at around 45 dBA idle, and 52.5 dBA at full load. Ambient temperature may impact these numbers slightly.

I’m not super familiar with sound engineering, but I thought dB and dBA were different measurements. Their specs list the idle @ 45dB and peak @ 52dB. Can anyone with more experience chime in?

Any tests I can run for you??

dB and dBA are technically different, but I’ve seen more people use them interchangeably than use them “correctly”. dBA just has the lower frequencies weighted to mean less sound since it is harder for humans to hear these lower frequencies:

Thanks Wendell,

I’m mostly looking for power usage and sound levels both have when filled up with enterprise drives. I know the HL15 could have some very powerful CPUs thrown in, but I was looking at the “Fully Built & Burned In” HL15.
If there’s some other motherboard, CPU, HBA combo that is confirmed compatible for running TrueNAS (SCALE or Core) that would be fairly low power (and similar price range) I’d be interested in throwing that in the HL15 too. Or if there is some other case/appliance option that meets the criteria I listed above.

I’m willing to take a hit to noise if power efficiency is that much better.

Looking at that curve I guess it’s possible fan noise is similar between dB and dBA?

yes, most definitely if between 1k-10khz which is where the more annoying fans usually fall.

I used an iPhone and the NOISH sound level app.
A foot away the existing server was 66dBA and when right in front of the server it was 70dBA. So it sounds like either option will be an improvement.

@Altkey Hello! Thanks for this thread. I’m contemplating the exact same choice right now, for the exact same use case.

As much as I’d prefer the HL15 for expansion slots and more drive space, I’m space constrained and the Mini R would fit better. Pricewise, I"m going to have to sell things/save up either way.

I didn’t realize until I looked at this thread how much power the HL15 (potentially) consumes vs. the Mini R, though I suspect that usage as little more than a file server would keep the HL15’s power draw lower.

What did you decide to do? How did it go?

1 Like

The plan is to purchase this server before March. I haven’t pulled the trigger yet though.

I’m currently leaning towards the TrueNAS Mini R because of the lower cost, lower power usage, smaller form factor, and official TrueNAS support.

I’m hoping someone with both can provide some feedback on actual noise levels and power usage so I can feel a little more confident.

Does the PCIe slot support bifurcation? I’m looking at an x16 NVME carrier board that says it needs bifurcation to work properly. (I’d like to use this box as a VM store, so more NVME would be helpful.)

I’ve confused myself quite a bit looking at photos of the inside of this unit.

I’m trying to figure out if the PCIe slot is open so it can take x8 or x16 cards.

Someone got a (fish-eyed) view of the internals.

I also found a picture of the (same?) board in a review of the Mini X+.


It looks open, but I’m not sure.

I’m pretty sure atleast the early TrueNAS Mini R’s use the A2SDi-H-TF motherboard which does not support bifurcation. That being said TrueNAS is famous for switching motherboard OEMs around while keeping the same system names so later TrueNAS Mini Rs might have Asrock Rack motherboards.

A PCIe x4 slot on motherboard can’t bifurcate to support a x8 or x16 nvme carrier card since it needs the lanes to physically match up to all the lanes of the installed PCIe card… unless you get a nvme carrier card with a PCIe switch built in (but those tend to be very expensive).