[solved] Can't claim a Plex server from separate vLan

[nihvel]

Qnap - Server Version#: 1.28.0.5999 (latest as of 12/Oct/2022)
CentOS Stream - Version: PlexMediaServer-1.29.0.6244-819d3678c (latest as of 12/Oct/2022)
Ubuntu 22.04 - Version: PlexMediaServer-1.29.0.6244-819d3678c (latest as of 12/Oct/2022)
Player Version#: N/A

I redeployed my whole home-lab using a Proxmox clusters of 3 nodes:
1x RPI 4B
1x Intel Nuc 8gen
1x Custom PC with AMD Ryzen 5600x

One of the storages is a QNAP TS-451DeU.

I used to run Plexmediaserver on the NUC on a Ubuntu 18.04 (10.0.2.13) and later on a Ubuntu 20.04 (10.0.2.14) VM, with ESXi as Hypervisor, and also as an app on the (now dismissed) QNAP TS-231P (10.0.1.150).
My Plex account was able to discover both of the servers. I didn’t have to do anything for this to happen.

On this new environment, only the Plex APP of the QNAP is listed in the Servers of Plex WebUI.
The other 2 VMs, Ubuntu and CentOS, even after logging in with the Plex account from their own private IP address, are not seen in the “authorized devices” section.

If I stop the Plex media server service on the QNAP, I am still able to login to Plex WebUI from any of the 2 VMs, using their own IP, but I am warned that “Plex02”, the QNAP Plex, isn’t online.

Qnap IP: 10.0.1.200
Ubuntu VM: 10.0.20.25
CentOS VM: 10.0.20.23

The QNAP (10.0.1.200) hosts the folders with my collection. I use CIFS to share those folder with the VMs.
Connection between all the devices involved work. There is no blocking rule yet between the 2 vLans.

Fun fact:
If i open http://10.0.20.23:32400 (CentOS. But same happens on Ubuntu) and try to add a new library (only Plex02 is shown, the plex qnap app), I can only see the Filesystem of the QNAP because of the Plex qnap version that is also running).
Both Centos and Ubuntu don’t show up as servers, even though I am logged in on all of them using the same Plex account, and yet I can edit Server (qnap) and player related settings.

I am not sure if this is a bug but there’s definitely an error somewhere.
Can somebody please help me?

[tom80H]

You should be able to claim those “missing” servers. Once they’re linked to your account you should see them from any app:

1. open the web app bundled with each server locally → using the server’s local IP address from within the same subnet (e.g. http://10.0.20.23:32400/web for your CentOS setup)
2. go to Settings > [Server Name] > General
3. click the Sign In button at the center of the settings page (next to the “this server is unclaimed…” message)

Edit: make sure to do this from within the same subnet; your servers seem to be spread across different subnets, so that might take some extra effort!

You probably did that at some point in your old setup or you had configured your old system to not require authentication when accessed from certain subnets…

[nihvel]

Hello!
No I don’t have Settings > [plex01].

To (re)start “clean” I removed all the devices from Authorized devices and also uninstalled the Plex App from the Qnap.

I re-installed Plexmediaserver both on CentOS and Ubuntu, but I’m lately only doing the initial config on Ubuntu.

• apt reinstall plexmediaserver
• open the page at http://10.0.20.25:32400/web/index.html
• I log in
• I am welcomed with the “select your streaming service” window, which i close
• I go to Settings and there is no server. There isn’t even the section for the server.

When I had the Qnap app installed, I did see it.

I’ve been using Plex since 2012, I won’t say I know it good, because I was never interested in deeply know it, but truth is I miss the manual server configuration and the possibility to skip the login.

Anyway, my previous setup had 2 LANs (2 eth cables) 10.0.1.0/24 and 10.0.2.0/24 with the servers on the subnet 2 and the storage on subnet 1. Pfsense in the middle between the 2.
No particular blocking rule between them.

My current setup has the same firewall but 1 cable with several vLans plus the tagged LAN. 10.0.1.0/24 remained for the Qnap, but will be moved soon.
10.0.20.0/24 where all the VMs and nodes are.
Here as well, there is no blocking rule. Actually, there’s no rule at all if not for the allowAll2All. I’m still configuring it.

I did not do anything special in the previous setup. Literally. The 2 Lans were 1 hop away, the firewall. This time, it is the same. Just different network.

Of course I will try to setup a VM in the 10.0.1.0 sudnet, and I will see if the issue is in the network.
Otherwise, if even there is a setting on Plex, like many other software have, where I can whitelist/accept connections from a list of subnets, that’d be the real deal!

I’ll leave here a screenshot, maybe it’ll helps

[nihvel]

I installed Plexmediaserver on my Fedora with dhcp address 10.0.1.24 and it works. Looks clear that it doesn’t like my vLans.
I don’t know why.

[nihvel]

There is definitely something wrong with Plex and how it runs under different subnets.

I decided to continue with CentOS just because it always is my first choice when I have to setup a new server.
I open a ssh tunnel and forwarded the port 32400 to my local Fedora. After the login I had the initial server setup wizard.

I hope this will help somebody in the future.

Don’t mind the ‘sssh’. That’s an alias.

[dane22]

@nihvel

10.0.1.0/24 10.0.20.0/24

The two above are on separate IP Nets

In order to claim a server, the workstation doing that must be on the same IP Net, and if not, I suggest an SSH tunnel

See this article, and search for On a Different Network: https://support.plex.tv/articles/200288586-installation/

[nihvel]

In order to claim a server, the workstation doing that must be on the same IP Net, and if not, I suggest an SSH tunnel

Yes, well, I kind of understood that (:

Thanks for linking the docs. I didn’t find this page while I was googling keywords like “plex on different vlan” and such.
Anyway, I’ve been running Plex on a separate subnet since 2017. This is the first time I fall into this “feature”.
Glad it works now.

Cheers,