client request timeout via VPN

[Putiput]

my jellyfin server runs just fine on a raspberry pi 4, i can access it locally and play media and access via a wireguard vpn running on the same pi works as well. Just updated and rebooted the whole thing today.

I'm trying to give a friend access via the same wireguard server, browsing works fine for him but when trying to play files he gets an error: "Playback Error There was an error processing the request. Please try again later"

I get some errors in the server log, I've attached the whole log, but the relevant part is probably this:

[2025-01-03 14:52:14.744 +00:00] [INF] WS "192.168.178.160" request
[2025-01-03 14:53:21.899 +00:00] [WRN] WS "172.18.0.2" error receiving data: "The remote party closed the WebSocket connection without completing the close handshake."
[2025-01-03 14:53:21.902 +00:00] [INF] WS "172.18.0.2" closed
[2025-01-03 14:53:46.475 +00:00] [WRN] WS "192.168.178.160" error receiving data: "The remote party closed the WebSocket connection without completing the close handshake."
[2025-01-03 14:53:46.475 +00:00] [WRN] WS "192.168.178.160" error receiving data: "The remote party closed the WebSocket connection without completing the close handshake."
[2025-01-03 14:53:46.475 +00:00] [INF] WS "192.168.178.160" closed
[2025-01-03 14:53:46.478 +00:00] [INF] WS "192.168.178.160" closed
[2025-01-03 14:54:03.529 +00:00] [ERR] Error processing request: "Reading the request body timed out due to data arriving too slowly. See MinRequestBodyDataRate". URL "POST" "/Items/13d4a4200ea009e4528c86acab2ec81b/PlaybackInfo".
[2025-01-03 14:54:16.402 +00:00] [INF] WS "192.168.178.160" closed
[2025-01-03 14:54:16.922 +00:00] [INF] WS "192.168.178.160" request
[2025-01-03 15:00:05.477 +00:00] [INF] Sending ForceKeepAlive message to 1 inactive WebSockets.
[2025-01-03 15:00:17.477 +00:00] [INF] Sending ForceKeepAlive message to 1 inactive WebSockets.
[2025-01-03 15:00:29.478 +00:00] [INF] Sending ForceKeepAlive message to 1 inactive WebSockets.
[2025-01-03 15:00:29.478 +00:00] [INF] Lost 1 WebSockets.

The client is running Linux mint 17 and jellyfin is installed via flatpak, I'm not sure where to find any clientside logs.

the wireguard config (if relevant) is:

[Interface]
PrivateKey = XXXXXXX
Address = XXXXXX

[Peer]
PublicKey = XXXXXX
PresharedKey = XXXXXX
AllowedIPs = 192.168.178.146/32
PersistentKeepalive = 0
Endpoint = XXXXXXXX

I've already tried setting persistentkeepalive to 1 and changing the config to route the entire traffic via the vpn, made no difference.
Not sure what the web socket error means to be honest, any help would be appreciated.
edit: also tried accessing via web browser, result was the same

[TheDreadPirate]

If the server is running on a RPi, why is it successfully transcoding via Intel QSV?

Code:

-codec:v:0 h264_qsv

And the architecture is reported as X64.

Code:

[2025-01-03 13:30:54.928 +00:00] [INF] Architecture: X64

And this looks like a plain deb package install and not a flatpak (which is good).

Code:

[2025-01-03 13:28:36.450 +00:00] [INF] Program data path: "/var/lib/jellyfin"
[2025-01-03 13:28:36.450 +00:00] [INF] Log directory path: "/var/log/jellyfin"
[2025-01-03 13:28:36.450 +00:00] [INF] Config directory path: "/etc/jellyfin"
[2025-01-03 13:28:36.450 +00:00] [INF] Cache path: "/var/cache/jellyfin"
[2025-01-03 13:28:36.450 +00:00] [INF] Temp directory path: "/tmp/jellyfin"
[2025-01-03 13:28:36.450 +00:00] [INF] Web resources path: "/usr/share/jellyfin/web"
[2025-01-03 13:28:36.451 +00:00] [INF] Application directory: "/usr/lib/jellyfin/bin/"

Having said all that, try changing the AllowedIPs to include the whole subnet.

Code:

AllowedIPs = 192.168.178.0/24

[Putiput]

You're absolutely right, sorry. I kind of forgot I upgraded the whole system a while ago to a mini PC. I could post the stats if that helps, but this issue seems to be only with one client, who has the issue on both his flatpak client and browser.

Changing the allowedIPs didn't help unfortunately, thanks though. Routing the whole traffic, as in changing AllowedIPs to the default 0.0.0.0 etc didn't help either.

Does the client save logs somewhere that could help?

[TheDreadPirate]

How far away is your friend from your server? Do you know if there is any packet loss between your server and them?

[Putiput]

my friend is about 600 km away, however i was able to reproduce the issue with my laptop from my parent's place, about 30 km. ping shows no packet loss:

Code:

ping -c 20 192.168.178.146                                          (base)
PING 192.168.178.146 (192.168.178.146) 56(84) bytes of data.
64 bytes from 192.168.178.146: icmp_seq=1 ttl=63 time=24.0 ms
64 bytes from 192.168.178.146: icmp_seq=2 ttl=63 time=18.1 ms
64 bytes from 192.168.178.146: icmp_seq=3 ttl=63 time=22.4 ms
64 bytes from 192.168.178.146: icmp_seq=4 ttl=63 time=20.5 ms
64 bytes from 192.168.178.146: icmp_seq=5 ttl=63 time=34.9 ms
64 bytes from 192.168.178.146: icmp_seq=6 ttl=63 time=20.4 ms
64 bytes from 192.168.178.146: icmp_seq=7 ttl=63 time=35.4 ms
64 bytes from 192.168.178.146: icmp_seq=8 ttl=63 time=20.5 ms
64 bytes from 192.168.178.146: icmp_seq=9 ttl=63 time=77.9 ms
64 bytes from 192.168.178.146: icmp_seq=10 ttl=63 time=22.2 ms
64 bytes from 192.168.178.146: icmp_seq=11 ttl=63 time=27.1 ms
64 bytes from 192.168.178.146: icmp_seq=12 ttl=63 time=17.1 ms
64 bytes from 192.168.178.146: icmp_seq=13 ttl=63 time=22.7 ms
64 bytes from 192.168.178.146: icmp_seq=14 ttl=63 time=16.8 ms
64 bytes from 192.168.178.146: icmp_seq=15 ttl=63 time=21.6 ms
64 bytes from 192.168.178.146: icmp_seq=16 ttl=63 time=20.7 ms
64 bytes from 192.168.178.146: icmp_seq=17 ttl=63 time=18.5 ms
64 bytes from 192.168.178.146: icmp_seq=18 ttl=63 time=20.6 ms
64 bytes from 192.168.178.146: icmp_seq=19 ttl=63 time=20.5 ms
64 bytes from 192.168.178.146: icmp_seq=20 ttl=63 time=20.5 ms

--- 192.168.178.146 ping statistics ---
20 packets transmitted, 20 received, 0% packet loss, time 19092ms
rtt min/avg/max/mdev = 16.846/25.125/77.874/13.034 ms

this is the output of traceroute:

Code:

traceroute to 192.168.178.146 (192.168.178.146), 30 hops max, 60 byte packets
1  * * *
2  192.168.178.146 (192.168.178.146)  19.824 ms  19.763 ms  19.759 ms

[TheDreadPirate]

What happens when you configure jellyfin to bind to a specific address?

Dashboard > Networking > Bind to local address > 192.168.178.146

[Putiput]

I've most likely resolved the issue, thanks for your help. it was an issue with DNS, adding the line DNS = 1.1.1.1 to [Interface] and adding 1.1.1.1/32 to AllowedIPS like this:

Code:

[Interface]
PrivateKey = XXXXXXXXXXXX
Address = 10.8.0.4/24
DNS = 1.1.1.1

[Peer]
PublicKey = XXXXXXXXXXXX
PresharedKey = XXXXXXXXXXXXX
AllowedIPs = 192.168.178.146/32, 1.1.1.1/32
PersistentKeepalive = 25
Endpoint = XXXXXXXXX

works for at least one client. the other one has issues starting wireguard with this config but that is unrelated to jellyfin.
thanks again