paperblankets

Yes, you have to be running Unraid Nvidia as well. .

Any reason that can't be a nested directory like `/tmp/plex` To keep this container from being able to crawl the tmp files of anything else running on the unraid server? Random people online claim: I guess I assume that will be true for docker engine on linux. I guess I will give it a shot and see what happens... Edit: That seems to work fine since docker runs as root.

Yikes, so the typical plex instructions of just using /tmp are not adequate? So you're saying it should be: With: Or are you mostly just saying, whatever you do, make sure you have a path mapping configured for TRANS_DIR?

Hi there, I'm attempting to set up this container for the first time. I am getting a port in use error (Starting the docker container) on `443` I believe because I am running letsencrypt. Error starting userland proxy: listen tcp 0.0.0.0:443: bind: address already in use. Edit: I just set an explicit https port the way I would with other containers and that seems to be working for serving the admin UI at least.

When you each say you followed the tutorial, does that include the framebuffer unbind as a user-script that runs with each each array start? Also note that quite a few of us have not been able to get a bios posted from TPU to post our GPUs, but editing the bios dumped form our actual card worked.

Each automated release has the plex media server build version listed (example, 1.19.2.2673): https://github.com/binhex/arch-plexpass/releases Copy that build version, search for it in the plex release announcements: https://forums.plex.tv/tag/release-announcements Bam, changelog. Plex does not have a releases page on github or anything (PMS is not public). Someone let me know if there's a more convenient way to do this, I'm curious as well.

I'm in the same boat. I have a functioning windows vm with a 2080 passed through. That same bios dump is not working in steam os.

Hi Acps, That's something you handle yourself, navigate to your appdata directory for this container: `./appdata/binhex-plexpass/Plex\ Media\ Server/Plug-ins/` (for me `/mnt/disks/ssd240/appdata/binhex-plexpass/Plex\ Media\ Server/Plug-ins/` since my containers run off my main array) There you will find any bundles installed for plex: 403897278 drwxrwxrwx 3 nobody users 61 Nov 2 2018 Audiobooks.bundle/ 403897285 drwxrwxrwx 3 nobody users 39 Nov 2 2018 HdbitsSubtitles.bundle/ 1339160 drwxrwxrwx 3 nobody users 22 Oct 10 18:41 Sub-Zero.bundle/ 1287830 drwxrwxrwx 3 nobody users 58 Nov 2 2018 Trakttv.bundle/ 270108051 drwxrwxrwx 4 nobody users 130 May 20 2019 WebTools.bundle/ So that's where you want to copy your new bundle. Good luck!

Some things they get access to, for instance, if you make a plex premium music library, anyone with access to your server has access to that. I believe some of the ui features are only for the user signed in with plex pass. The plex forums will probably be more knowledgeable about which features are available to which user types on a premium server. A list of features I think are shared: Library types, trailers. Features I think are not shared but could be wrong: offline, audio leveling, dvr/recording, tidal,

What user are you sshed as? The exact script I run is: ``` #!/bin/bash echo 0 > /sys/class/vtconsole/vtcon0/bind echo 0 > /sys/class/vtconsole/vtcon1/bind echo efi-framebuffer.0 > /sys/bus/platform/drivers/efi-framebuffer/unbind ``` Only one vtcon path is valid at a time for me, so I frequently get a `tmp/user.scripts/tmpScripts/Fix mmap/script: line 3: /sys/class/vtconsole/vtcon1/bind: No such file or directory` Error, but I leave both so my script keeps working if I change video cards or pcie slots.

You're likely to run into the msi issue next (It will manifest as poor performance, and distorted audio sometimes after driver install). Just follow the instructions to msi enable your drivers and that issue should stop after a reboot of the vm ( ).

Idk what to say there. I see 2019-11-22 17:50 was the last package time (well after when I made my comment if I'm handling timezones properly), but I pulled on that repo (https://aur.archlinux.org/plex-media-server-plexpass.git) before posting and I had gotten updates that I _swear_ were a new release. Either way though I'm not under pressure to get this release on my server. I was just trying to illustrate to @Barbadus where he would look if he wanted to take the container update into his own hands.

I would not try it, but I believe you can ssh into the container, source or reinstall aur ( https://github.com/stuartpb/aur.sh ), which seems to have the latest plex pass version: https://aur.archlinux.org/packages/plex-media-server-plexpass ). I'm sure you can upgrade or force install the package. I just don't hotfix my containers...high effort, high risk of doing something not intended.

I'm eagerly awaiting 6.8 stable, I think I'm going to hold off on the RC's. I ran into the corruption issues a few months ago and rolled back. Your method did work for getting the install media to boot. My install issues seemed to be related to the install failing to properly format the drive. Eventually I just saved onto the partition of a previous failed attempt and that worked.

Is this still the best way? I'm finding the installer stalls during install, and eventually restarts. I'm deleting my vdisk and starting fresh. In case something with vdisk I created is the issue. Unraid 6.6.7 (Awaiting 6.8.* to avoid database corruption issues) Settings: Base VM: Arch Distro: I've tried both Manjaro KDE and XFCE Primary vDisk bus: Virtio Graphics: VNC Cirrus Bios: OVMF Machine: Q35-3.0 Both installers have been stalling at or before 30%. I've tried all the above with `systemd.mask=mhwd-live.service` set as well. Edit: I got a stable install by wiping my entire hard drive from the install media, and then selecting the ext4 partition I wanted to install onto.

I do have to warn you, you will have serious io limitations running plex on a usb stick. Make sure to use a usb 3.0+ port and drive, anything less will just be a mess. I don't have any great evidence or articles to back up the io limitations but the gist I hear is usb is much more cpu bound than sata or something else that offloads work (I would love to have some solid reading on this from someone who knows more though). You can absolutely use a usb stick to run the plex appdata folder (via the unassigned devices plugin) In my case, the ssd my appdata is on has a single partition called appdata. The process will be the same for you, except on a usb stick.

In the past I have had my watch history synced to trakt, and I would just pull it back down if my server got into a bad state. But at this point the only client that I have that can still interface with plex plugins is the Android TV client on my shield. So I expect any time I will no longer be able to restore my watch history. I have no solution to that problem at this time beyond good backups (I only keep one backup nightly and it's burned me so many times, I need a better system as well). For sure

Yes it (restoring the backup) should resolve it if your database was in a good state at the time of the backup. I can't speak to if 6.8.0-rc1 resolves the issue, @Squid said above he thinks it may help.

You do have one other option if you have the disk space, you can use unbalance to move files off of one of the disks in your array, and then remove one of the 6 drives from the array. You can then use the unassigned devices plugin to mount one of your drives next to your unraid array, and you can run appdata off of that disk. This is what I currently do in unraid, my docker images are not part of my parity drive...but a nightly backup copies appdata to my unraid array.

Some day I will update my current fans, I have a ton of supermicro stock PWM fans and the noise is real! I hope you enjoy the new rig, honestly, the last few upgrades I have done the best part has just been knowing my gitlab server is not running on hardware I feel might be unstable/fail any time. I guess peace of mind has become more important to me than raw performance.🤗

What did you upgrade to?

Ha, by chance, I got the exact error you got last night. In my case, one of two things occurred: A windows vm started that was allocating the same gpu as my plex container. This shut down the plex container. (Any attempts to start it result in `libusb_init failed` in the logs). I did some cable management in my living room and disconnected and reconnected some usb hubs, somehow this caused issues with my plex container. My assumption is #1 was the cause. You may not be on nvidia-unraid though, so maybe this was some other issue. Simply stopping my offending vm in my case allowed my plex container to start again. The solution to #1 was simply to stop the offending vm, no restart was required. Hopefully part of that helps you debug your issue!

There is probably a better way, but I just copy my entire appdata/binhex-plexpass directory (I also back it up nightly in case something goes wrong) I'm also interested to know if there is a better way. You can backup the plex database by itself, but that would not help you with your server configuration, thats what I don't want to setup again in my case. Other thoughts: If your appdata directory is on a cache drive...make sure to stop the container, and invoke the mover before you copy the directory or you will end up in an old state/be missing the changes on the cache drive. I just don't keep my containers on the cache drive. It's more trouble than it's worth for me.

If you look at your unraid logs around that time, do you see any errors/failures reported?

My super lazy way to do it, is add the appdata directory as a share for a moment: Then I can just copy the files over from whatever machine is convenient (While the plex container is stopped). You could run into some permissions issues, but those can be resolved from a shell without to much difficulty. Edit: Just remember to remove the appdata share when you are done...you don't want that shared/public most likely.