Comment on What's the security situation when opening a jellyfin server up for casting?

• dogs0n@sh.itjust.works ⁨1⁩ ⁨week⁩ ago

  If you don’t want to worry too much, you can setup a vpn (like wireguard) on your server for ssh access.

  Using a non standard port is a good idea, but not entirely foolproof because bots might still port scan (even if unlikely that they do that for ssh I’m not sure). At a mininum, you probably want to use keys for login like the other commenter on the main comment said.

  Personally, using a vpn for when I want access to SSH when I’m out is worth the couple hours setting it up the one time (very simple setup with wireguard-easy for example). Maintenence time spent on upgrading is very low.

  (Tl;dr I’d use a vpn to access ssh specifically rather than exposing it to the internet)

  Same thing for Jellyfin?

  Not 100% sure what you mean, but to clarify: Don’t accidentally expose jellyfins port to the internet (eg the default port 8096). Make sure it is only accessible from outside your network through your reverse proxy.

  source

  • victorz@lemmy.world ⁨1⁩ ⁨week⁩ ago

    Could you explain a bit more?

    Like, right now, I have two machines in my local network. Both are running sshd on port 22.

    In my router, I’ve set the port forwarding to be some high port number in the 19000’s to forward to port 22 on the first machine, and then the same high port number incremented by one (1) to forward to port 22 on the second machine.

    Also key based login only of course.

    Is this insecure in some way?

    Would a VPN make connecting to my computers more secure somehow? I’m not sure I understand how if so.

    ---

    What I meant with the Jellyfin question was kind of, how is having it exposed via a reverse proxy different from exposing its port right away? It’s it because the only allowed connection would be HTTPS/encrypted etc, maybe?

    I’ve never set up a home network apart from physical cables and using routers and switches before, no advanced site/network configuring. Definitely interested to learn more though for when I want to serve a real media center using a NAS and like a Pi.

    source

    • dogs0n@sh.itjust.works ⁨1⁩ ⁨week⁩ ago

      Your SSH setup is good.

      ssh is a very resilient piece of software so I doubt with your setup you would encounter any issues.

      Enforcing use of a VPN to get into your network before being able to ssh into a machine is mostly just an extra layer of defense, though using a non-standard port, only allowing key logins and disabling root user login are all layers of defense you have already added.

      I thinj you’ll be fine, but if you are worried, you could setup a VPN or alternatively something like Fail2Ban if you notice any brute-force attacks (which may be unlikely with the use of a non-standard port).

      What I meant with the Jellyfin question was kind of, how is having it exposed via a reverse proxy different from exposing its port right away? Is it because the only allowed connection would be HTTPS/encrypted etc, maybe?

      It’s down to how secure the software is really.

      Jellyfins (and other software) don’t use really secure web servers for getting themselves accessible via the network.

      Caddy (a reverse proxy, for example) is made to be exposed to the internet and so it is more resilient and safe to use.

      So putting the resilient software (a good reverse proxy) infront of Jellyfin (or most other software) simply increases your security by having the more safe web server be the one interfacing with end users.

      Have fun on your journey!

      source

      • victorz@lemmy.world ⁨1⁩ ⁨week⁩ ago

        Okay cool, thanks for that follow-up and confirming my SHH setup seems reasonable. 🙏

        There’s one thing I don’t really get though, with the whole reverse proxy thing and how that’s supposedly safer:

        putting the resilient software (a good reverse proxy) infront of Jellyfin (or most other software) simply increases your security by having the more safe web server be the one interfacing with end users.

        Like, once a user client has contact with the Jellyfin instance, via the reverse proxy, wouldn’t the Jellyfin instance be just as vulnerable as without the reverse proxy? Once a connection is established, or found to be available, you could just start exploiting away in the same way, right? Or wrong? If wrong, how? 😅 Maybe it’s too long for a text reply? Maybe I should watch some helpful video explaining how it works. 😁

        source

        • -> View More Comments