search

ressu / kube-plex

Scalable Plex Media Server on Kubernetes -- dispatch transcode jobs as pods on your cluster!
Apache License 2.0
104 stars 23 forks source link

support: no errors, but video doesn't play #17

Closed ohthehugemanatee closed 3 years ago

ohthehugemanatee commented 3 years ago

Got it running with an NFS-mounted /transcode directory. Transcode pod is started correctly, and /transcode fills up as the transcoder runs. No problems reported in the console. But... the video is stuck on buffering at 0:00.

What is the desired flow, here? Console suggests that the transcoded segments are sent back to the master with a PUT request. Is that right? Or does the master just read the shared /transcode directory? Any advice on debugging further would be appreciated.

kubePlex:
  image:
    repository: ghcr.io/ressu/kube-plex
    tag: latest
    pullPolicy: Always

timezone: Europe/Berlin

ingress:
  enabled: true
  hosts:
    - plex.vert
    - plex.example.com

nodeSelector:
  beta.kubernetes.io/arch: amd64

persistence:
  transcode:
    enabled: true
    accessMode: ReadWriteMany
  data:
    claimName: "nfs-claim"
    subPath: "media"
ressu commented 3 years ago

Plex reads the files directly from /transcode while the transcode process sends status updates back to the main process via a POST request (I think it was POST). If you are seeing dark screen only, the Plex process is most likely unable to see the files on the shared directory.

Looking at your configuration, I'm not quite sure what your transcode PVC looks like. I think the default in the template is an emptyDir which isn't shared between processes, so that could be the cause...