Plex "No soup for you!"

[pinoli]

Hello,
I have been searching for this specific problem in the TrueNAS SCALE forum but I could not find anything.
I found some reference about the same problem for TrueNAS CORE that was not helpful.

I am running TrueNAS SCALE Angelfish 21.01 ALPHA on the hardware in signature (ask if you need more info).
I tried the official Plex app, provided the claim token, mounted a SMB share to /media folder, assigned the Intel GPU for HW accelaration and finally deployed.
I can access to the web portal by using the PORTAL button on the deployed app in the Installed Apps section of TrueNAS, I can login into Plex and authorize the new server (plex.tv shows the local IP of the server and asks me if I want to authorize the connection).

At this point I am greeted with the infamous
No soup for you! The server you're trying to access doesn't want to let you in. Make sure you're signed in as a user with access to this server.

I am kind of lost, since the suggestion I have seen here does not apply to me, because I don't have the Network section in my Plex web UI.
Also, I connected to the pod using the Shell feature of apps in TrueNas SCALE, but I can't seem to find where Plex settings are located.
I found binaries in /usr/lib/plexmediaserver and service scripts in /var/run/s6/services/plex, but preferences.xml is nowhere to be found.

Thanks in advance for any help.

 

[ornias]

21.01 ALPHA doesn't exist, do you mean 21.02 ALPHA.1?

 

[ornias]

First off:
It's helpfull to provide actual screenshots of the settings you entered while installing an App.
It's almost impossible to bugtrace things when you write a story about something.

"mounted a SMB share to /media folder"
I can already tell you the permissions are wrong and hence it isn't mounted. How do I know? I spend 2 days fixing the mount-permissions issue for TrueCharts :P

"The server you're trying to access doesn't want to let you in. Make sure you're signed in as a user with access to this server."
Because your storage might have wrong permissions, preferences.xml might not be saved correctly. Hence on restart you create a new-fresh plex server. However the claim token is only valid for 4 minutes. So: No soup for you.

That being said:
I cant know all of this for sure without:
a. screenshots of the settings you entered
b. ls -l readout, on the host, for the folders you are trying to mount

 

[G8One2]

"No soup for you" - Get your PC on the same subnet the plex jail is on, you wont have that problem......

 

[ornias]

"No soup for you" - Get your PC on the same subnet the plex jail is on, you wont have that problem......

This is about SCALE not about CORE
You can't do this with k8s, thats what the claimtoken is for.
Claimtoken adds the server to your account, which makes sure you don't need the same-subnet-auth-bypass.

 

[SweetAndLow]

This is about SCALE not about CORE
You can't do this with k8s, thats what the claimtoken is for.
Claimtoken adds the server to your account, which makes sure you don't need the same-subnet-auth-bypass.

Ssh port forward will get it on the same network. Just an idea to help get it working.

 

[Samuel Tai]

See the bottom of the official Plex docker repo for the SSH port forward instructions if you forgot to apply the PLEX_CLAIM on container creation.

GitHub - plexinc/pms-docker: Plex Media Server Docker repo, for all your PMS docker needs.

Plex Media Server Docker repo, for all your PMS docker needs. - plexinc/pms-docker

github.com

 

[ornias]

See the bottom of the official Plex docker repo for the SSH port forward instructions if you forgot to apply the PLEX_CLAIM on container creation.

GitHub - plexinc/pms-docker: Plex Media Server Docker repo, for all your PMS docker needs.

Plex Media Server Docker repo, for all your PMS docker needs. - plexinc/pms-docker

github.com

@SweetAndLow and @Samuel Tai , there are more/other issues at play here.... Don't skip over the storage issues Both I and OP highlighted, there is a connection between the two. If his permissions are off, no solution around the plex_claim is going to solve this.

It's also important NOT to point users on SCALE to all sorts of hacks. This can get perfectly solved within the SCALE eco-system and within the SCALE-APP without ssh port forwards and such.

The correct way of solving this on SCALE is:
- Fix and verify storage
- enter new claim token
- Done

It's the easiest way of solving this. Period.

 

[SweetAndLow]

@SweetAndLow and @Samuel Tai , there are more/other issues at play here.... Don't skip over the storage issues Both I and OP highlighted, there is a connection between the two. If his permissions are off, no solution around the plex_claim is going to solve this.

It's also important NOT to point users on SCALE to all sorts of hacks. This can get perfectly solved within the SCALE eco-system and within the SCALE-APP without ssh port forwards and such.

The correct way of solving this on SCALE is:
- Fix and verify storage
- enter new claim token
- Done

It's the easiest way of solving this. Period.

Looking forward to reading your pull request.

Until then my workaround helps someone get it working.

 

[ornias]

Looking forward to reading your pull request.

Until then my workaround helps someone get it working.

I think you don't understand how this is (primarily) a storage issue. The storage issue can be solved, which allows him to hit the big "edit" button, enter a new claim_token and get it running again.

Actually the complete steps would be:
- Remove App
- Fix and verify storage permissions for hostPaths (default is root:root for ix_volumes, should also work for your own mountpoints using hostPath)
- reinstall app using new claim token
- Done

Your solution doesn't solve it, because the user (seems to?) report his plex doesn't save the preferences correctly and SCALE Apps are NOT persistant by design (unless storage is mounted). So it would break again on any restart of the k8s pod.

And my fix for the storage issues for TrueCharts has already been merged, a few days ago.
Thats TrueCharts, not official...

When I say something might involves a very specific SCALE bug and your solution is not going to fix the issue (on its own), i'm not saying that to annoy you, but because I actually had to fix this storage issue for 30+ Apps (which also lead to preferences not being saved correctly when using hostpaths with "wrong" permissions).... just like... OP (seems to report)

*edit*
That being said:
This can "easily" be fully prevented by adding a "add host network to allow list" flag and spinning up a pre-install,pre-update job that adds the host IP range to the list of IP's not requiring authentication.
Sadly enough due to a bug I currently am unable to work on the TrueCharts plex chart and I don't have the time to also work on official charts (besides already spending time bugtesting the Apps system in general)

 

[pinoli]

I think you don't understand how this is (primarily) a storage issue. The storage issue can be solved, which allows him to hit the big "edit" button, enter a new claim_token and get it running again.

Actually the complete steps would be:
- Remove App
- Fix and verify storage permissions for hostPaths (default is root:root for ix_volumes, should also work for your own mountpoints using hostPath)
- reinstall app using new claim token
- Done

Your solution doesn't solve it, because the user (seems to?) report his plex doesn't save the preferences correctly and SCALE Apps are NOT persistant by design (unless storage is mounted). So it would break again on any restart of the k8s pod.

And my fix for the storage issues for TrueCharts has already been merged, a few days ago.
Thats TrueCharts, not official...

When I say something might involves a very specific SCALE bug and your solution is not going to fix the issue (on its own), i'm not saying that to annoy you, but because I actually had to fix this storage issue for 30+ Apps (which also lead to preferences not being saved correctly when using hostpaths with "wrong" permissions).... just like... OP (seems to report)

*edit*
That being said:
This can "easily" be fully prevented by adding a "add host network to allow list" flag and spinning up a pre-install,pre-update job that adds the host IP range to the list of IP's not requiring authentication.
Sadly enough due to a bug I currently am unable to work on the TrueCharts plex chart and I don't have the time to also work on official charts (besides already spending time bugtesting the Apps system in general)

Well, I cannot thank you enough for all the great insights.
Regarding the OS version, yes it was a typo, I am sorry. I also had it in the signature, so honest mistake.
If you say that permissions could be a problem, then I need to go back to work, since not everything is still clear to me. Also bear in mind this is my first experience with TrueNAS at all (no FreeNAS, no TrueNAS CORE), so bear with me if I am not precise.

To test if it works without first setting shares and permissions, is it acceptable to run PMS with no storage attached? I ask because I am quite sure it did not work even when I did not mount any shares. Claim token was always fresh.

Anyway, I will give it a couple more tries and report back.
Thanks

 

[ornias]

To test if it works without first setting shares and permissions, is it acceptable to run PMS with no storage attached? I ask because I am quite sure it did not work even when I did not mount any shares. Claim token was always fresh.

If you do not check "Enable Host Path for*insert name* Volume" it will auto create what we call an "ix_volume" for you, that's attached storage thats created and managed by SCALE. It is also automatically reverted if you ever need to rever an update of the App.

So I would suggest:
- Removing the App
- Installing with a new claim_token, but not checking any of those boxes mentioned above
- verify the installation works
- Click "edit" on the installed App
- and check the "hostPath" checkbox for ONLY the "data" volume. That means you would only need to manage the permissions for that single volume, where "data" means "your media files"

That "should work"(tm)
But be mindfull: claim_tokens only last 4 minutes max.

 

[pinoli]

I can report back on the first 3 bullets.

Unfortunately I cannot make it work.

• I completely removed the app, created a new one with an untouched configuration (except for the claim token)
• Previously removed any authorized device from my Plex account to start clean.
• I did not even attach the available Intel P630 for HW acceleration.
• I then logged into the Plex portal immediately after the pod had been deployed (well within the 4 minutes window), only to get the same sad message.

I did not even try with shares because I am still getting my head around it, but as far as I understand it should work out of the box with the default configuration. Yet, it doesn't.

Before doing anything else I am going to reinstall TrueNAS SCALE from scratch, to eliminate any potential post-installation mistake I may have made in some configuration: Plex will be the first and only thing I will run after a clean install, and I'll see if the problem is still there.

 

[ornias]

Dont reinstall scale, thats silly. The chance of scale being the fault is VERY low.

Share things like logs and errors.
What i’m primarily interested in is the content of /mnt/pool/ix_applications/releases/plexname/volumes/ix_volumes

And how you got your claim key

Another option would be:
- Execute the solution above (from @SweetAndLow and @Samuel Tai )
- claim the server
- check if claim worked by connecting normally
- Stopping the app
- (re)Starting the app.


If that leads to reintroducing the error, that means storage is the issue.