I’m having an issue with my server where new devices (that haven’t connected to the server before) are failing to find My Server.
Symptoms
Even when using a different browser on the same Computer that is able to connect fine on Chrome’s session and the Windows Plex media server session fails to find the server. This is especially odd because even after remote tunneling into the server and going to localhost:32400 it gives a “A server could not be found.”
Steps taken to Resolve
I was able to get a temporary workaround where, after removing the tokens to every device (deleting all attached devices to my account), It was able to then see the server as unclaimed. After claiming it, the new device (ie. Different Browser tunneled to localhost:32400) was able to communicate with the server.
However, This fix lasted only several minutes. After re-sharing with all my shared users, and checking on my phone (having to re-log in as I had removed the token prior), now the android app cannot seem to find the server. Same with My Samsung Television, and a huge chunk of my shared users.
LOGS
Attached are some Javascript Console Logs from an Incognito Chrome Window Failing to find the server. Plex - Logs.txt (18.5 KB)
You can see what appears to be the main error here: "Stratos is unauthorized at [insert location here] (Status 401)
I have Filtered out my domain, WAN IP, and the active plex token (Which should be noted as it did not correspond to the one that I had recorded earlier after reclaiming the server)
You can see it is also trying to communicate with the server “Stratos” at the local IP (It is on the same LAN as the computer attempting the connection) 172.16.0.3
Though I’m not sure where it is pulling 69.164.196.66 as this is not My IP and geolocate does not show it near me.
Server OS: Ubuntu 18.04.1
Server Version#: 1.14.1.5488
Though this issue has been present since the latest 3 releases as far as I’m aware.
I mean, if you go through the logs, you will find the lines with “Stratos is unauthorized at” and then an attempted location.
There is the local IP that it attempts to connect at: 172-16-0-3
There is the Domain it attempts to connect at: Filtered to MYSERVERDOMAIN.TLD
And There is the WAN IP: Filtered to MY-PUBLIC-WAN-IP
All of which show unauthorized no matter which network I connect from on a device that cannot access (Local and WAN)
This is happening to (seemingly) random devices on any account that I share with.
Wondering if on the server you may have to go in and clean out your registered playback devices, that are registered to your server. Then reboot your server, then have the devices connect to it again, to re-establish the browser / device list to the PMS to correct it.
I have done this already multiple times.
The first couple devices to connect can see the server and content, but after that, every device after fails to find/connect to the server
How do you reckon?
I’ve also tried disabling options related to custom certificate domain, and custom certificates, and disabling NGINX. After all that, the issue still persists.
I’ve disabled the firewall on the Ubuntu Server during testing, no results.
I’ve removed all attached devices, and re-tunneled into the local network and claimed the server again. After doing so, the device in which I am connected can see the server. If I connect any new devices (for a time) It can see the server as well.
By the time I re-add all the user shares, new devices are, once-again, failing to connect to the server.
Arte you using an SSL cert?
Do you have more than one active against your public IP?
ie, plex has a different cert than something else you have running like SONARR or w./e
Edit: what are you using for router/firewall and what are your nats, rules.
How are you handling DNS if at all?
I am using the same cert that I have been using for nginx for the single domain on the server.
Plex is running on an extension to that domain and not a subdomain through a reverse proxy through nginx.
E.g. domain.tld/web/
But, as I said, even disabling certificate settings and stopping nginx, the issue still goes on.
My server is on a DMZ to the web through a single-NAT setup. Thus, port forwarding is a non-issue. My server connects to google’s DNS for looking up records, and my domain has the proper A-records to my server’s WAN IP.
create a new install under a new folder name, manually configure it. kill the other server executable and test the new one, if the issue goes away migrate your library and users.
if it does not go away it’s related the ubuntu server
I just did that like a half hour ago. Only things that have been reimported from the old install were the Database and Metadata.
Another thing of note.
I am using the Plex-API (via the plexapi python library) to re-share with all connected users after re-claiming the server. Doing this by hand takes a lot of time. I do get a new token every time by grabbing the one found from inspecting an XML file on the server.
To denounce that as a flag of interest though, I have done this process before, and re-added users by hand, without touching the API, and the issue persists.
I have not messed with that, would be outside my wheelhouse.
you are likely going to need to share some plex or server logs to get further along with the troubleshooting. Sounds like you have introduced several points of possible failure, you will want to rule them out one by one.
