Yes, that’s exactly what I’ve done. You still haven’t shown me why it’s unsafe. If you can’t, that’s fine. At some threshold or another, nothing is secure. The one thing I know for sure is that that first page, that says it doesn’t recommend exposing a port, does not say what you said it does.

Probably the number one recommendation I see in self hosting communities is to not open ports directly (other than for a reverse proxy). It seems like a common recommendation no matter the service. To be clear: I am a beginner. I know very little about this, but I’ve spent months learning. I can’t say you’re wrong, but I don’t think you’ve made a convincing argument for me to actually understand why Jellyfin is unsafe to expose to the internet compared to any other service.

It says exposing a port directly to the internet is not recommended; do you know of any project that would recommend directly exposing a port? What is meaningfully different here?

By all means correct me if you know more, but what I tend to see is one or two people here saying that Jellyfin devs don’t recommend exposing it publicly, only to be corrected by looking at the actual documentation. I suspect those cautioning against it are on outdated information and that Jellyfin carries much the same risk as exposing any other service.

I’m currently only hosting Jellyfin (movies and TV) and Komga (books and comics) on a LAN, but I already bought a domain, and I’ll be exposing them via reverse proxy within the next few weeks; maybe even this weekend if everything goes well. After that, Nextcloud is a must for sharing large files, as Google Drive’s limits are coming up on me fast, and I’ll start test driving Fluxer as a replacement for Discord.

I don’t think I can read books fast enough to make automation attractive either, haha. I break DRM on Calibre with two plugins, books purchased from Kobo, and then organize on Komga after manually dragging the files.

Correct. I have received help from that server in the past, and it’s unusable now. Part of what I hope to accomplish is to replace Discord, or at least offer my friends an off-ramp. Getting started with Discord is way easier, and in the year 2026, I wouldn’t know what other client to use to look for a community that gives me quick help via live chat. I no longer know how to use IRC, I’ve found.

Hey, just wanted to chime in and say thank you. I think your guide moves a little fast for someone like me, but through omission, I was able to suss out what was wrong, I think. I don’t know if it was a default setting or if it was something I picked up without understanding it while trying to fill in the gaps of DNSmasque DHCP, but I had two DHCP Options set; one was a Set option for router[3], and the other was a Set option for dns-server[6]. The fact that you didn’t have that in your guide at all led me to try a configuration without them, and now I’ve got full connectivity on my VLAN. I’ll of course now start properly blocking access off rather than leaving everything totally permissive before opening up services to the web.

When I was given a similar suggestion, I asked why I would need proxmox for my project, and I was told something along the lines of, “Don’t discount their usefulness if you’re trying to do this as a career.” I am not trying to do this as a career. I already have one of those. I’m trying to replace subscription services with something more economical and under my control.

I don’t suppose your issue looks anything like my thread from this past weekend in this here community? And as a reminder, tagged ports face networking hardware, while untagged ports face end devices.

Oof, I feel you OP. I started down this road back in February, and I thought I’d be set up by now, but I’m still learning some of the pieces for a project that’s almost exactly the same as yours. If it’s any consolation, I have made a ton of progress. The hardest part can often be when you’ve chosen the pieces you want to use for your project, and then a kind stranger who means well adds extra complexity above and beyond what you want or need on top of their suggestion.

Hey, that’s forward progress! The first I’ve had since this thread! That command did in fact allow me to ping my desktop from my mini PC on the VLAN. It also allowed me to ping 8.8.8.8, which I was unable to do before. On reboot, that default gateway seems to be reset until I run the command again, which makes sense. So I guess my next question is: what does this mean, and how do I fix it? I take it to mean that of those three jobs DHCP is supposed to provide in your list, it’s only done the first one. The DHCP systems in particular are a major change from what OPNsense was just two years ago when the guides I’ve been following were made.

So what do you think might be the problem with it that I might check? As far as I know, if I’ve been assigned an IP address, DHCP is working.

I won’t be able to try this until at least tomorrow after work, more than 24 hours from now, but yes, this was what was in my tutorials as standard. But also agreed in those tutorials was that what I tried doing was even more permissive and ought to “just work”, at the risk of security and with no separation; I always try to do “Hello World” before I try to implement anything of value, and right now I’m barely seeing “Hello”.

They’re on different subnets, so I figured. I believe DHCP is set up properly, because it assigned an IP to that mini PC in my specified range.

Thanks, I’m going to have to go through this with a fine-toothed comb to see where I went wrong. If you read my other comments and have a hunch as to the problem in the meantime, I’m all ears, lol.

The output of that command is:

It was listening when I pinged the gateway from the mini PC.

I might not understand what you mean, but doing the best I can figure out, the output from the endpoint mini PC running ip route is:

192.168.10.0/24 dev enxc84d4422aa48 proto kernel scope link src 192.168.10.157 metric 100

From the OPNsense firewall, the trace route looks like this (I would have expected to see the switches that it hops to in between, but I don’t see them here):

I can’t find any option to print routing tables in my switches, both of them Netgear GS305E switches. I don’t see any mention of it in the manual either. I suspect that what you asked me to do was lost on me.

I watched and rewatched tutorial videos, traced my cables and ports, and configured the port settings until I was able to ping the gateway on this VLAN. That’s as far as I’ve gotten. Over the course of today, since posting this thread, I’ve rechecked those settings a handful of times, and they still appear to be correct as far as I know. If I delete the extremely permissive firewall rule that I set up for the VLAN, I lose the ability to ping the gateway, which seemingly validates the rest of my setup and leads me to believe that this is a configuration issue in OPNsense rather than the configuration of my switches…but I don’t know what I don’t know, and I’m still learning this stuff.

I understand that you’re recommending what you think is best based on your experience, but as I’ve been trying to learn self hosting with a semi-simple goal in mind, the extra complexity that folks keep recommending around just about every facet, because their needs or desires are greater than mine once they’re more seasoned than me, does make it all more difficult to take in during the learning process. Maybe I’ll want to go more advanced some day, but for now, the goal is to host fewer than a dozen services off of two different devices that live under my office desk and consume under 100W between them. I want VLANs for this as a means of separation in case the security of my exposed services is compromised, but with this smoke test, I want to prove that I understand the basics of doing so, so it’s currently feeling defeating that I don’t. I don’t want to sound like I’m not appreciative of any help you can offer, but I do still believe that simpler is better for me at this point.

My firewall mini PC has four ports, but only two of them are active; LAN and WAN. I got that much working without much fuss and replaced my ISP’s provided router. There were two dumb switches between the firewall and the office, but once that was working, I replaced them with managed switches; when they’re not yet configured, they’re indistinguishable from dumb switches. I’ve been over my OPNsense configuration a dozen times in this thread by now, but let’s just say this new VLAN is set to be as permissible as I know how to make it, coming very close to my default LAN interface settings as far as rules go. They ought to be identical. The two smart switches are set up such that port 5 is “in” and 1 is “out”. Living room 5 connects to the firewall. Living room 1 connects to the office switch’s port 5. Office switch 1 connects to the end point mini PC. Living room ports 1-5 are untagged for VLAN 1; ports 1 and 5 are tagged for VLAN 10. Office ports 2-5 are untagged for VLAN 1; for VLAN 10, 1 is untagged and 5 is tagged, and port 1 has a PVID of 10.

I spelled all of that out in hopes that I did something stupid that I don’t know how to spot but maybe you do. Every device on VLAN 1 is working as it should with internet access. The one device on VLAN 10 only has access to the gateway and nothing else, despite the most permissive “allow everything” rule I could set up.

That’s what I’m saying though. I believe I tried a few configurations of both 4 and 6 now, in the time since you asked. At first it was only 4, because a tutorial said I didn’t need to bother. I added one for 6. I tried setting the source to “any” instead of just the VLAN network. I tried a handful of permutations of those things, disconnecting and reconnecting the wired connection to release/renew between each one. It doesn’t seem to have changed.