Using the default setting for the local relay in the Helm chart for k8s results in the transcoder failing. In the plex-worker logs, it looks like it is getting a 502 Bad Gateway when it tries to post the chunks back to the relay.
To Reproduce
Steps to reproduce the behavior:
Deploy Cluster Plex with Helm and Remote Only Workers
Try to transcode
Expected behavior
Transcoder should work as expected.
Screenshots
plex-worker log
[http @ 0x7f9b9426be40] HTTP error 502 Bad Gateway
[tcp @ 0x7f9b9e14ee80] Starting connection attempt to 10.101.77.235 port 32499
[tcp @ 0x7f9b9e14ee80] Successfully connected to 10.101.77.235 port 32499
Describe the bug
Using the default setting for the local relay in the Helm chart for k8s results in the transcoder failing. In the
plex-worker
logs, it looks like it is getting a 502 Bad Gateway when it tries to post the chunks back to the relay.To Reproduce Steps to reproduce the behavior:
Expected behavior
Transcoder should work as expected.
Screenshots
plex-worker log
nginx error log for relay
nginx access log for relay
Desktop (please complete the following information):
(not a client issue)
Additional context
Cluster is on bare metal running Talos OS 1.7.6 and k8s 1.30.3. PMS is 1.40.5. Cluster Plex is 1.4.13 (latest everything, just installed today).
Disabling the relay, whitelisting all IPs and adding the plex-pms service IP to the
pmsIP
in the Helm chart fixes everything.