2024-08-29, 10:38 PM
Hello!
I am running Jellyfin Server 10.8.13 on Windows 10, the server has 3 IP addresses (from the same subnet) bound to the same IPv4 TCP/IP interface (.150, .160 and .161).
I am using a letsencrypt certificate (from a public domain I own) with Jellyfin inside my LAN for local clients to connect using https://htpc.mydomain.whatever
Both IPv4 and IPv6 configured and are functional in my LAN, however for Jellyfin I am only using a "A" record (IPv4).
The communication between clients on the LAN and the Jellyfin server are working without a hitch (It has been a good while since I have setup all this), but I always configured the clients manually (using the server FQDN matching the DNS A record plus the default port 8920 for https)
In my Jellyfin server, the setup which is different from the defaults (to my understanding) in terms of networking would be:
General, Server name: htpc.mydomain.whatever
Networking, HTTPS Settings: "Require HTTPS" checked
From a new and never configured before jellyfin client (it's the android one for this particular test), when I click on "Choose Server" I get presented with a single choice containing the information below (2 lines):
htpc.mydomain.whatever
https://x.x.x.150:8920
I see the response to discovery does tell what my server name is, but for some reason that FQDN is not being used for the connection string.
For obvious reasons, attempting to go ahead nevertheless and connect to https://serverIP:8920 fails.
I vaguely remember (when I first set up Jellyfin originally) being able to use the provided response for a http connection. (That was before including a TLS certificate and forcing use of HTTPS in the server configuration, naturally).
I captured the response UDP packet with wireshark running in the server to further confirm what I saw in client:
Any suggestions on what I could tweak in the config for the server auto discovery to return the address https://htpc.mydomain.whatever:8920 as a response to client broadcasts?
Thanks10.8.13GeneG
I am running Jellyfin Server 10.8.13 on Windows 10, the server has 3 IP addresses (from the same subnet) bound to the same IPv4 TCP/IP interface (.150, .160 and .161).
I am using a letsencrypt certificate (from a public domain I own) with Jellyfin inside my LAN for local clients to connect using https://htpc.mydomain.whatever
Both IPv4 and IPv6 configured and are functional in my LAN, however for Jellyfin I am only using a "A" record (IPv4).
The communication between clients on the LAN and the Jellyfin server are working without a hitch (It has been a good while since I have setup all this), but I always configured the clients manually (using the server FQDN matching the DNS A record plus the default port 8920 for https)
In my Jellyfin server, the setup which is different from the defaults (to my understanding) in terms of networking would be:
General, Server name: htpc.mydomain.whatever
Networking, HTTPS Settings: "Require HTTPS" checked
From a new and never configured before jellyfin client (it's the android one for this particular test), when I click on "Choose Server" I get presented with a single choice containing the information below (2 lines):
htpc.mydomain.whatever
https://x.x.x.150:8920
I see the response to discovery does tell what my server name is, but for some reason that FQDN is not being used for the connection string.
For obvious reasons, attempting to go ahead nevertheless and connect to https://serverIP:8920 fails.
I vaguely remember (when I first set up Jellyfin originally) being able to use the provided response for a http connection. (That was before including a TLS certificate and forcing use of HTTPS in the server configuration, naturally).
I captured the response UDP packet with wireshark running in the server to further confirm what I saw in client:
Any suggestions on what I could tweak in the config for the server auto discovery to return the address https://htpc.mydomain.whatever:8920 as a response to client broadcasts?
Thanks10.8.13GeneG
