Can you show me in the TOS (https://www.plex.tv/about/privacy-legal/plex-terms-of-service/) where you see that?
In the TOS, I see no text indicating that I cannot create a Plex cluster on my own servers, at home, for local use.
Plex’s logic is a bit strange if what you’re saying is true, because you can have multiple Plex servers, but you can’t create a Plex cluster.
I don’t know if this applies to you, but if you use the automatic database backup feature, the directory where the db backups are written must be owned by plex or tmp files will appear there alongside the backups.
You are right, but this is not my case, as the following configuration ensures that all volumes declared for the container have the same rights and the same user.
runAsUser: 36548
runAsGroup: 36548
fsGroup: 36548
fsGroupChangePolicy: OnRootMismatch
supplementalGroups: [36548]
I apologize for sounding harsh so will clarify my statement.
What has happened in the past,
I personally have 52,000+ media files in my library.
I can stream to all 8 players, plus my 9 shared family members, simultaneously from ONE server.
While not expressly stated in the ToS, the maximum number of shared users is 50.
Will all always be streaming from your server(s) such that you need multiple?
This is what’s not making sense to me,
If this is for local use, given you didn’t say how many servers, how many servers do you need ?
I believe there is a misunderstanding:
To date, and for about 2 years now, I’ve only had one Plex server in my HomeLab:
In two years, I haven’t had any operational issues with Plex, so much so that I didn’t even pay attention to it, apart from version updates.
Until yesterday, I was using replicated storage with Erasure-Coding, and now I’ve switched to simple replicated storage (undergoing validation).
The advantage of using Plex with replicated storage is that when I perform maintenance on one of the cluster K8S nodes, Plex restarts in less than 15-20 seconds on another node. And the big icing on the cake is that it’s completely transparent, and no one notices that the server has been restarted. No one complains in the house, which makes me the happiest person in the world. I can even restart Plex twice in a row without anyone noticing anything.
While trying to update the version of Plex, I noticed that the storage was taking up over 500 gigabytes. Initially, I thought it was logs and that there must be a problem with them. But upon further investigation, I discovered that it was due to the database, hence the creation of this ticket.
Then, you explain the cause of the problem. I thank you for that.
But to fix my problem, you tell me that I need to switch to local storage in a K8S cluster.
If I switch to local storage in a K8S cluster, I lose volume replication across nodes, and consequently, I lose my 15-second restart and create a service interruption during the maintenance of the node where Plex is located. Whereas before, I had zero stress performing maintenance on my K8S cluster; I could do whatever I wanted, whenever I wanted, however I wanted. The node could remain off for a week, it bothered no one. By moving Plex to local storage, node maintenance management is no longer the same.
All the happiness acquired until then disappears at once.
The question I ask myself is how, with local storage, can I get what I had before?
One of the solutions is to deploy 2 to 3 Plex instances in the cluster. However, this solution comes with disadvantages. This is because I would consume three times more resources (CPU, RAM) for Plex, and a solution needs to be found to synchronize the three Plex instances since I want it to be completely transparent, as before.
Hence my question, is it possible to synchronize and have a clustered Plex under the same claim?
So in the end, I have nothing commercial in my management. It’s just that local storage poses a big constraint for me.
I am closing the topic, I no longer have any problems with -tmp files
