Leaderboard

We are extremely close to releasing HexOS Local, our solution for accessing/managing your HexOS servers through a locally installed application on your server. Before we roll this out to all current users, we are looking for a handful that are able/willing to help us test the migration process. The requirements are pretty simple: 1) Users that agree to enroll in this early access test will have ALL of their servers migrated (we cannot apply the test to just a single server, so if you have multiple, it's all or nothing). 2) Users must have a discord account as those accepted will be invited to our new discord server for the purposes of bug reporting and support needs. If you are interested in helping us test this, please drop a DM to @mill3000 with the subject "HexOS Local Early Access" and be sure to include your Discord username in the message. Any questions, post them here! Thanks everyone!!

I'm out of town right now but let me just state that I totally get the folks that are upset. Honestly, writing blogs is pretty stressful for me as we constantly are working on moving targets and I stress over communicating about things that are still in flux and getting the words right. Tbh, I stress over comms in general. It's nerve wrecking. It's why I am earmarking funds this year to bring someone in to handle this for us. I just didn't want to spend any money on non dev/support resources until we got to 1.0. Again, I get it and I am not gonna make any excuses for it, so if you want to be mad, please direct that anger at me, not my team. They are busting their butts to get this thing polished and rolled out.

Another month, huh? The least you could do is respect the people waiting with some communication ahead of time. Instead it's silence all day on release day and a last-minute delay notice as if it wasn't clear days ago this could happen. You created this self-imposed deadline,, nobody forced it. A delay with advance communication - fine. But the silence until 9 PM? That's just pure disrespect

I don't know if "many" are thinking that, @vogam7, but I'm not. As @G-M0N3Y-2503 says this is par for the course in software development. "Communication" is a lot more work and headache than you would think and their time is much better spent on making the software.

So just to be clear... The initial deadline (end of 2025) got delayed. The self imposed new deadline (Q1 2026) got delayed. The blog post about the delay got delayed (and still not out). The extent of the communication in the meantime on they delay is a random comment buried in a random forum thread. Radio silence on the official comms (reddit, twitter, newsletter). Hard to see how that could be considered acceptable / consistent / reliable. What's the level of goodwill here to assure that even end of April release will be met? The track record so far now is the opposite - the self imposed deadline means nothing.

https://docs.hexos.com/blog/2026-04-02.html

Likely unpopular opinion, but thank you!

This one is entirely on me. I am traveling today but I will be working on getting this blog post up tomorrow (April 2nd). I'm also planning to hire someone to take over marketing/comms this year as a full time position.

It also could be the case that they've done a bunch of internal testing and are just gradually rolling it out. But In software development delays occur, and people will be upset. It's way better to phrase an initial role out as early access testing in case there is an unforeseen showstopper. Than say it's good and then be called out for buggy software. So really it's a matter of if they hire PR or a community manager or not. I vote that there are better ways to spend that money. Considering that roadmaps and timelines are just a guess to attempt to make people happy, from my perspective they've been about as inaccurate as I'd expect, so no red flags from my perspective.

Should be no risk to data loss at all. The biggest risk is getting disconnected from our deck for server administration. That's why we want users to be able to talk to us via discord so that we can quickly remediate those types of problems if they occur. The big switch here initially is going to be authentication. We're moving to key cloak and we set up a more robust infrastructure to enable us to migrate users between production and development environments very easily. While we've done a ton of internal testing on this, we just want to be extra cautious before doing a mass migration of thousands of users all at once. Technically this is still pre 1.0, but the only things that really remain for us are hexos local and some improvements to our health monitoring systems.

The teams working as hard as they possibly could on this and when participating in a beta delays, bugs and broken updates are to be expected. Personally I would prefer delays over broken updates.

1.0 should be out today. When will it be pushed out?

The nuance I see is: Guesstimate for initial release, end of 2025. Oh shit, people want a local UI? Guess that's fair but that's a tone of extra work... but here's hoping... (I like many assumed the local UI was a part of the original plan) End of 2025: So it turns out the local UI was a tone of extra work, but we have a better guesstimate, Q1 2026. 1 Week before the end of Q1: Ok beta-beta release, If this goes flawlessly we might just make it... End of Q1: So transitioning X all at once is a massive migration that you can't really test for that well, so we'll have to do it gradually so we can cater for the inevitable fallout. Maybe we can do that over a month? Obviously I'm completely guessing (see, lack of communication) and could be completely wrong, but the general rule I've seen is go off the magnitude of the guesstimate, the longer the period the less accurate it is.

Working on a blog post with an update. We had to push back to April due to some issues with our health system that we've been tracking down since February. The plan is to both begin the rollout of HexOS Local and complete the health system upgrades and then release 1.0 by end of April. I was hoping to get a blog up today with all of this information, but it just didn't come together fast enough.

I messaged the user. Thanks for the shoutout!

@jonp @mill3000

Wow, Communication is a headache? That's a laugh. Well it looks like you have come to the right place. You only get very filtered communication when things are going really good here. And for anyone who actually was a developer you already know communication is critical. A lack of communication leaves you with a forum like this. Dozens of questions that get ignored. When did communication and coding become mutually exclusive? Right now the only people talking positive about HexOS are a few folks that depend on this product for their livelihood. Fortunately most people seem to not care or never took anything said up to this that serious. And looking at how many issues there still are unanswered in the forums and the length some of those have been around. Some issues have been able to celebrate birthdays. And if you want something to think about, riddle me this. If communication is like this now when we are in "BETA" and there is still goodwill stemming all the way back to the LTT video, What will it be like when we have HexOS 1.* and the product is release to production. Do you think the communication will get better then? The entire team has made it easy to get down on this product/project but if your going to be honest with yourself you still see the possibilities that attracted you in the first place. If you love the product and you are honest with yourself you see the cracks and holes that give you an uneasy feeling. Everyone in the middle is pretty meh in that it will do what they need but maybe not what they want. And just off on the horizon is the real issue we have been dancing around. A perfect HexOS is still just a skin wrapped around parts of TrueNAS. It will never be more than that and no matter how hard you look you were never promised more than that. If you don't want to learn TrueNAS then welcome to a sturdy and easy to setup network attached storage appliance. As Immich has shown us multiple times, the application curation process was just to make the initial install easier and more standardized. It is not to make Immich work or eliminate code product issues and design changes. What kind of communication do you expect when a new release of an app or perhaps TrueNAS itself wants to change something fundamental like their storage or security models. And from who? It's probably time we stop worrying about when production version 1.0 is coming and start trying to figure out what we are willing to do to make it work. Learn TrueNAS? Buy new hardware? Buy consulting services? Cut your losses and find something else? Regardless of all the rest , it seems you can't count on communication or timelines here. If you can't live with that no matter what you thought in your mind this journey would be, well sorry because this is just the start.

@Ammonia some people seem to be getting that to work in this thread

As the title said, I have been holding out since I purchased the key. Recently heard on the LTT podcast that 1.0 was launching at the end of March. Is that still the plan? I haven't seen an official blog update in a few months.

I did say BY end of April ;-). Truth is we have been pretty much right there aside from a few issues, but the big one was the availability of HexOS Local. But we decided yesterday that we can still call this 1.0 and roll out Local to everyone over time. The local project has to be rolled out more slowly as we have SSL cert issuance limits right now.

I have just recently added a To GPU to the NAS to help with Plex, would be nice to have statistics for it on the main dashboard page. Also might be nice to have a settings page that allowed you to select to use it.

Love this guys content!

Don't get me wrong, the communication is at what i would call a minimum, any less and I'd guess it was abandoned. But, it's a consistent and reliable minimum so far. I check the forum every other day, so would i appreciate more details, definitely. But would i trade anything for that outcome, probably not. So with the minimum communication and knowledge of the software development process patterns, I'm content enough. But that's not an expectation for others I have either. If you'd trade a license for that outcome and the team decide that they should improve communication because of it, hey, that's good too,

The people saying everything is fine have got to be software developers who see this only from one perspective. Hate to break it to them, most people we'd expect soon to pay full license fee to keep the project going aren't, and those are exactly the potential customers we'd lose with this laid back mindset to weak areas of this project. Those won't be beta expectations anymore and there will be a lot to prove to doubters of this project to justify the price. Imho plex & immich was genuinely exciting to watch grow naturally, and backing them early felt like a bargain especially towards stable launch. Was it perfect? No, there were delays too. But none I recall having such communication meltdowns close to stable release. I'll be honest, some of the responses here concern me as much as the delays and comm issues themselves. Some say I knew it's part of the process "when I read Q1 2026 I think Q3 2026" maybe, but not this far into this beta with no updated written delay comms. Others react confused when I try to hold the team to their own deadlines. 'I'm glad they're not communicating, that means time is much better spent on making the software...' really? Good teams do both and I'm not asking for daily extensive blog posts of what and why went wrong, just a sentence update across all hexos communication channels setting new expectations. I recall people on this very forum calling out poor communication back in July and October, and the team acknowledging it and promising to improve. Yet here we are. Traveling or not, I still don't understand how the whole team forgot the most anticipated day of their own project, or couldn't break the news of a delay at least the night before. Now whether April Fools or not, yesterday would've been the first day after stable launch with customers paying $300 for a license, needless to say the most expensive nas os on the market. At this point, if the team still offers a refund option, I'd happily take it.. this just isn't for me anymore

Well @vogam7did you get what you wanted? If you have read these forums there are more than enough accusations of terrible or completely lacking communication. I have been jousting with this windmill since shortly after the initial deployment. I've even talked about to many methods of communication that left each incomplete. I got a similar response it would be looked into. and I can say this was the only answer you could have expected. They are probably still a small team primarily of coders. There had been talk months ago of no marketing team, no support team, and only initial talks of ongoing funding. If we are thinking this communication level is the biggest issue we are going to face before this project stabilizes, we are seriously kidding ourselves. That LTT video is more of a curse than blessing because this project isn't ready for that kind of attention. This team isn't going to acknowledge they are having trouble. They just won't say anything, again. And there is nothing any of the paying beta customers can do about it. Watch and wait or don't watch, what's the difference now. It's to late to change the approach the team took, even if they have found that. And there is now reason to assume that's true. Did they take the right path to this development? Did they build the right team with the right leadership? Well, if you have a better plan then share it. If you don't like this approach how would you fix it? Communication alone means very little. Missing deadlines is a much bigger deal that not mentioning there are problems. And honestly, do you think this would be simple and smooth? Yeah right.

According to this picture found at https://nascompares.com/2024/10/04/jonsbo-n5-review/ , it seems like he has added the 4 SATA cables and 1 SATA Power cable to the board, no MOLEX. My best bet is that you can either use MOLEX or SATA Power, but don't need both.

"Working on a blog post with an update. We had to push back to April due to some issues with our health system that we've been tracking down since February. The plan is to both begin the rollout of HexOS Local and complete the health system upgrades and then release 1.0 by end of April. I was hoping to get a blog up today with all of this information, but it just didn't come together fast enough." -JonP "I am traveling today but I will be working on getting this blog post up tomorrow (April 2nd). I'm also planning to hire someone to take over marketing/comms this year as a full time position." -Jonp Link to discussion

Sorry if this has been mentioned already. There needs to be a way to store a boot disk config or recovery file on one of the arrays to allow easy recover of a dead boot drive (all done within HexOS! I shouldn't have to go into truenas if I don't know what I'm doing). I did install on redundant boot drives but bought too cheap and they both failed simultaneously. The only way I could find to recover was a fresh install without the array drives connected, then plug them in and manually recreate the exact share folder names in the HexOS from the list in truenas. Also, I would love to see a way to transfer or transition the boot drive to another drive by simply connecting a new drive. That way if all you had for a boot drive was a 1 Tb but wanted to use that for something else later you could get a smaller drive and transition.

It works

I'll just say what many are thinking here.. you're launching a hexos local beta just 9 days before the promised stable release date. How could this be included in 1.0 and called stable in under a week of testing? Refunds seem to be the go-to response for anyone who expresses frustration in other threads.. but that's beside the point for those of us who actually want this project to succeed. Missed deadlines are frustrating enough, but the communication issues that still keep coming up alongside them make it worse.. and for a project that depends on people investing in hardware, that trust matters. Server memory and now drives have nearly doubled since the originally promised end of 2025 release date. Every delay has a cost beyond just frustration. A straight timeline update would go a long way. I don't think you meant it this way, but the timing of this post is pretty telling... are we in for another delay?

At this point, just using HexOS to create folders, users, install apps, etc. just like normal. Use it both locally and remotely. That's really it. At this point, the main thing we're testing is that it works correctly, navigating you to the local interface when on LAN, and giving you remote access via the deck when away. If you can break your system (non-critical data only), there are a few specific cases we would like to see users test like destroying the pool used by apps (Docker) manually via TrueNAS, then renavigate to deck.hexos.com to see if you can remediate the issue. Obviously not asking anyone with real data on their systems to do this, but for those that have test setups, just beat the heck out it and see what breaks. We test internally pretty rigorously, but there's always edge cases that we can miss and this is the chance to find those as best we can.

if im not mistaken it looks like it isn't a docker application. you'll likely be able to get it running inside of a linux vm

I'm having the exact same issue on 25.10.2.1 — so the update alone doesn't fix it. I dug into it and found the actual root cause: The ix-vendor.service runs /usr/bin/start_vendor_service on boot, which calls vendor_service.py. This script starts a transient systemd unit called websocat.service (the websocket bridge between your local middleware and deck.hexos.com). The problem is that the script doesn't check whether websocat is already running before calling systemd-run --unit=websocat. If websocat is already active, the call fails with "Unit websocat.service was already loaded", but the script always exits with code 0 (it has a finally: sys.exit(0) block). Something then re-triggers ix-vendor, it fails again, and you get a rapid restart loop. Each loop iteration spawns a Python process that opens Docker sockets, and since the middlewared service only has a soft file descriptor limit of 1024, the FDs get exhausted quickly — hence the "Too many open files" error. Here's the fix that worked for me in truenas shell — a systemd override that adds a simple check before running the script: sudo mkdir -p /etc/systemd/system/ix-vendor.service.d/ sudo tee /etc/systemd/system/ix-vendor.service.d/no-loop.conf << 'EOF' [Service] ExecStart= ExecStart=/bin/bash -c 'if systemctl is-active --quiet websocat.service; then echo "websocat already running, skipping"; exit 0; fi; /usr/bin/start_vendor_service' EOF sudo systemctl daemon-reload sudo systemctl restart ix-vendor.service After this, ix-vendor shows active (exited) with status 0, websocat keeps running, and deck.hexos.com works fine. No more loop, no more FD exhaustion. I also raised the middleware FD limit as an extra safety net: sudo mkdir -p /etc/systemd/system/middlewared.service.d/ sudo tee /etc/systemd/system/middlewared.service.d/fd-limit.conf << 'EOF' [Service] LimitNOFILE=65536 EOF sudo systemctl daemon-reload sudo systemctl restart middlewared Both overrides live in /etc/systemd/system/ which - as far as I understand it "so far" - survives reboots. After a HexOS update you may want to verify they're still in place. Hopefully the HexOS team can add the is-active check to vendor_service.py itself so this gets fixed upstream.

If you're seeing 'Failed to start transient service unit: Unit websocat.service was already loaded or has a fragment file', then you can try: sudo systemctl restart websocat

Try the following. Most likely this will fix your issue. If this does, TrueNAS has a fix that is in it's next update. Login to TrueNAS interface. Left side select "system". Then after screen refreshes select "shell" from the list. Paste in the following command. *Can't use normal short cuts* Will need to use Shift + Insert Key for pasting. sudo python /usr/lib/python3/dist-packages/middlewared/scripts/vendor_service.py If that works within about 1 min you should be able to go to deck.hexos.com and see your server or claim it. You might need to run this after reboots. Let us know if this works. Thank you,

I... I was... I would have.... FUUUUUUUU-- Ordinarily, I would absolutely have tried another USB, and I actually was trying a bit ago, but that drive was being funky and wouldn't connect to Windows well and I gave up. I completely overlooked trying Balena Etcher (Rufus has always been rock solid for me so I never bother thinking aoub it) but NOW THAT YOU'VE SAID IT I feel like I remember needing to use it for CORE back in the day TOO! I am an IDIOT! Anyway, I grabbed a third USB, Etcher, and a new copy of the .iso, and set it all up again, and wouldn't you know it, the damn thing worked. Installed HexOS to the SSD, plugged in the HDDs and went to boot, and the box picked up the data drives to boot to. I thought I wasn't out of the weeds yet, but on a whim I tried re-enabling the B120i and going in and deleting the RAID array on the drives but keeping the single "array" on the SSD, which allows me to set the logical array as bootable in the B120i. Rebooted and... LOOK! HexOS!! I am as much at fault as any part of the technology in this thread, but in the end, through all that crap, and with your help, we got there, so... THANK YOU very much!!

Hey, please keep us posted about the progress. Concerning why you shouldn't use a HW Raid in combination with Truenas, the official ZFS documentation has a great section about this you find it here: https://openzfs.github.io/openzfs-docs/Performance and Tuning/Hardware.html#hardware-raid-controllers

Hello, It seems the after the update some systems on reboot are reaching out to our server so it doesn't know your there. Try the following. Login to TrueNAS interface. Left side select "system". Then after screen refreshes select "shell" from the list. Paste in the following command. Can't use normal short cuts Will need to use Shift + Insert Key for pasting. sudo python /usr/lib/python3/dist-packages/middlewared/scripts/vendor_service.py If that works within about 1 min you should be able to go to deck.hexos.com and see your server or claim it.

I got this working by doing a manual install of the docker compose file found here: https://docs.coolercontrol.org/installation/docker.html To manually install it, in TrueNAS go to Apps, Discover Apps, then click the three dots in the top right corner and load Yaml. You can leave the compose as is and it will work. You need to follow the guide below to setup the hardware for it. Specifically for TrueNAS that is running this command in shell and following the prompts: sudo sensors-detect To setup everything prior to that follow the guide here: https://docs.coolercontrol.org/hardware-support.html

Hi i just got Hexos up and running with a few apps and wanted to start getting my media server up and running. I installed jellyfinn and the Arr apps and got gluetun setup through dockge. I think now i just need to add gluetun into the networking in the .yml file for each arr app. For the life of me i cant find the .yml files to edit as they are made through hexos curated apps. Any way to access them to add gluetun for networking?

Hey, yes unfortunately this seems to be normal for BX APC UPS. Those UPS are not the best ones and the driver does not work very well with them. The problem seems to be that the USB part of the UPS keeps disconnecting and reconnecting every couple of seconds leading to this messages. It's a known problem. My advice would be, if you can still return it, do so and try looking for an Eaton UPS or similar. https://forums.truenas.com/t/back-ups-bx1600m-false-battery-replacement-alerts/50688

Offline Server for when SHTF happens. Anyone seen this yet? It’s from the guys over at crosstalk solutions. Wondering if it can be loaded as an app or a VM???

Dentro de la interfas de truenas, debes seleccionar la seccion de aplicaciones en la columna izquierda, dentro de la interfas de aplicaciones puedes ver todas las aplicaciones qie tienes instaladas, seleccionas la de imich y sale listado el puerto dentro de "workloads" o incluso puedes hacerle clic al boton "web ui", espero que te sirva!

It has gone rather quiet. I thought I would see more general discussion or the odd update. Unless I've missed something the most recent update has been a revision to the blog introducing hex OS local originally published 25th Nov 2025 and updated 23rd Jan earlier this year. I'm not complaining before anyone jumps on me. I've been happily running the beta since originally hearing about it on LTT and couldn't have built my NAS without it (not with my skillset at the time anyway). So far it's been running flawlessly, shares and apps have been a breeze to set up, and the minor jobs I've had to do myself such as setting up a VM in truenas for Homeassistant have been recognised and appear on the Hex OS dashboard. I know it's possible I might have to start again from scratch but fingers crossed there is an update path that doesn't break everything. I have on occasion had to re-adopt the server on the dashboard (probably my fault) but it's only happened a couple of times. To avoid wiping everything I just popped out my drives, re-adopted, shut down, popped the drives back in and everything came back when I restarted. If there isn't a simple update to 1.0 with any luck a similar workaround will prevent data loss.

i have this issue as well..march 2026 and still not fixed...i give up

Hello, One way that could fix this is to do the following. Update to TrueNAS version 25.10.2.1. which applies permanent resource limit fixes. Thank you

The reason Tailscale gets recommended so often is because it (and VPNs in general) is the most secure ways to access your services remotely. It's also pretty easy to install and get running. Nginx, on the other hand, is a completely different approach. To make lives a bit easier, let's take Nginx Proxy Manager instead of plain Nginx. Installing it is the easy part. The real work starts afterwards. You’ll usually need a domain, set up a DNS token for the certificate DNS challenge, configure something to keep your public IP updated, and create and manage all your subdomains. And all of that just to make your services reachable from outside your network. Once you’ve done that, your services are basically exposed to the entire internet. That means any device online can try to reach them. Because of that, you also need to put additional security measures in place to protect them from people trying to access or exploit them. Therefore only expose what you absolutely need to expose.

I had the same issue where HEX Deck would not recognize my server (deck said it was offline) but TrueNas UI worked. I realized it was my fault, I blocked it on my ubiquiti firewall accidentally when i was troubleshooting another issue. That's why truenas works because the HEX OS Deck has to call to Eshtek's server and the truenas ui does not. to answer another question in this forum, the local ui will mask this issue, not fix, because it will allow local connection, but wont change the fact that your server is blocked from external internet firewall level so connected to Eshtek wont work (only for this exact issue though). Another symptom of this exact issue was truenas did not have external internet access either, the app catalog was failing to update. Good luck to you guys hopefully this helps.

Happy to see that it's working now for you. 🙂 I still don't understand why this makes a difference which software you are using to create the USB drive, but hey at least it's running now. =D

I will be having family use my server for backing stuff up and as we all know end users always end up accidently deleting things and then wanting them back. Luckily, TrueNAS has a recycle bin option and hopefully HexOS can add it in one day as a setting that can just be turned on. For anyone who finds this and wants to do it, here are the steps: Set up the folder in the HexOS UI and then in the TrueNAS UI go to: Shares > Click edit on the SMB you want to add the recycle bin too > Purpose = No Presets > Click Advanced Options > Scroll to Other Options > Check Export Recycle Bin Then you will want to set up a job to automatically empty your recycle bin: Navigate to Chron Jobs in the TrueNas UI (just search for it but it's under System > Advanced Settings). Click add then you can either enter this into the command field as ONE LINE or save it to a file find /mnt/HDDs/FOLDER/.recycle/* -atime +30 -delete; find /mnt/HDDs/FOLDER/.recycle/* -depth -type d -empty -delete; If you want to test to see what files will be deleted with this command run it without the "-delete" If you save it to a file (like I did) you will enter this as the command. sh /mnt/HDDs/PATH_TO_FILE/empty_recycle_bin.sh Then just fill out the description, run as truenas_admin, and how often you want it to run and you are good to go. Hope this helps and hopefully it can just be a 1 click type of setup in the future!! P.S. Here is what the commands do: The first command searches for files and directories in /mnt/HDDs/FOLDER/.recycle/ that have not been accessed in the last 30 days and deletes them. The second command searches for empty directories in /mnt/HDDs/FOLDER/.recycle/ and deletes them. Also commands were stolen from here