clusterplex: Problem reaching plex from workers
Describe the bug It is giving me the following error from the worker container:
[tcp @ 0x7fcf32b59640] Failed to resolve hostname plex: Try again
To Reproduce Steps to reproduce the behavior:
- Go to plex main server
- Click on any video
- Check worker container log
- See error
Expected behavior It does the transcoding
Additional context
I am using a machine (low power) with plex + clusterplex_dockermod and also clusterplex_orchestrator. I have commented out port 1900 (it causes conflict when starting the container)
Another machine (high power) with plex + clusterplex_worker_dockermod.
The config is copied from the main server to the worker, at least for now, to make it easier to configure the /transcode path.
Main machine has a shared folder which worker connects to using cifs for data and transcode folders.
I can add whatever needed info here. What could be the issue? I see no reference to plex as it is in the configuration files
About this issue
- Original URL
- State: closed
- Created a year ago
- Comments: 26 (11 by maintainers)
Hi 4mazon, There’s more than one way to achieve this, but on the worker, you could add this to the docker file. extra_hosts: - “plex:ip address of plex instance” Example: extra_hosts: - “plex:10.10.10.1”
Cheers, Auzi
Ah, that’s unfortunate 😦 If I understood correctly, you are using CIFS, so not actually NFS, but Samba instead. Now, I haven’t deployed it that way in quite a while, but back then I had the mount on the underlying OS, and then just a bind mount to the container. The mount options I used back then where:
cache=loose,rw,vers=3.0,uid=1000,gid=1000,username=<user>,password=<pwd>I can’t remember exactly what the issues were, but I remember that specifyingvers=3.0was important for the share to the NAS to work.Absolutely, but it’s always good to get it working then work backwards from there.
You might need to try by forcing Plex and the transcoder onto the same host. I think this error is related to a directory missing (which might be transcode)
If you do the same set up but don’t use NFS (using
/tmp/transcode:/transcodeor similar), does it write successfully? It likely won’t actually play because plex would have to be on the same host, but does it write correctly?Have you tried setting
PUIDandPGIDenv vars to1000to match your NFS share? Or if it’s already set removing it?I’m just guessing, as I’ve said previously I’m not using NFS. However, this is definitely a permission issue with the share.
@4mazon in that video you’re performing actions as
root(0:0) and not asabc(65534:65534). If you do anls -laon transcode dir you’ll probably find 65534 doesn’t have permission to write to the directory. You can solve this by runningchown abc:abc -R /transcode/Transcodeorchmod -R 777 /transcode/Transcodeand then try again.You may however bump into permission errors with NFS since you’re setting the u:g to 1000:1000 in your NFS settings.