MediaMaan

Further to my last, I performed a reboot via SSH. unRAID then proceeded to boot normally, permit me to login, and is showing my drives and shares as usual. This time I used the GUI to create a diagnostic zip file, that looks quite different to the last. See attached. I'm really confused as to what is going on right now. unraid-diagnostics-20260706-1359.zip

OK, used Mint GUI to connect via SSH and ZIP the directory. Looks like most things are blank to me (e.g. no shares etc). unraid-diagnostics-20260706-1253.zip

Says it created a diagnostic file, but it isn't visible. I have found 3 diagnostic directories created in the root directory. Under /unraid-diagnostics-20260706-1343 I see 7 subfolders and 1 unraid-7.3.1.txt file. Not sure how to turn that into a ZIP file and move it from unRAID to my linux desktop here. No significant changes made at all. No new containers / plugins etc.

Hey all. Looking for some tips here. hoping I haven't been hacked. I run regular updates, and am on the current version of unRAID. However 3 days ago, I was unable to log in. The GUI was completely unresponsive. Couldn't SSH in either - kept getting connection refused. Had to hard power the server down :-( Came up fine though, and everything seemed good. Today, 2 days later, GUI was not working properly. Basically hardly anything being displayed on screen. Again, couldn't SSH in. Accessed terminal from connected keyboard and monitor, logged in and entered "reboot" on cli. Saw an error in /etc/initscript: line 22 /etc/rc.d/rc.6: input/output error. Hard to force power off again for second time in 3 days. unRAID rebooted, I logged in, figured I would check SSH is enabled (it was). Cancelled the Parity Check (as that would take days and had just completed 4 or 5 days ago. Saw there was a plugin update for Tailscale, started it, then saw an error the update was unavailable. Clicked Docker to check for updates, then got thrown out to a First-Time-Boot type of screen. I've been running unRAID Plus for years, and mostly just do updates. But something seems to have gone haywire. Every time I hit "Skip Setup" it just returned to this screen. After 5 minutes, I just tried it again, and the GUI is no longer accessible. [OK - figured out it was using https which is no longer valid, so can see the login screen at http://IP address. BUT, I cannot log in as root, just clears my password and returns to login screen] However I have now been successfully able to SSH in. I've suddenly got lots of questions! What is going wrong and where do I begin to diagnose / fix it? My first concern would be data preservation. How do I stop the array cleanly from an SSH connection? How can I restart the GUI? How can I diagnose what problems there are (e.g. is it my USB flash drive going bad, memory errors)? Has all my config totally disappeared? Extra information - my shares are currently inaccessible. I have a flash drive backup that is approximately 2 months old (from version 7.2.4) although unRAID had been updated to 7.3.1 (if that's what current is). Any help with next steps would be greatly appreciated! Thanks.

Ah - I am not running a DNS server at home. I'm using a third-party Secure DNS service. Although I expect it might be possible to create entries using that service, that could point to TS IP addresses. I generally only hear bad things about Plex. Especially relating to stopping people getting access to their own media on their own servers. So I am quite hesitant to make the switch. I actually only use Jellyfin because it does a better job with Profiles & Sharing across devices the Kodi. Kodi had been my goto for 10+ years. Played everything I threw at it, without needing to transcode. But, times are changing, and I am using way more Android devices these days (compared to devices flashed with LibreElec / CoreElec back in the day). This way I can access the myriad of other streaming platforms. I believe Plex has a very similar backend to Jellyfin, since it came first. But that means transcoding etc as well. If Plex could provide a bloat-free player that could handle the various codecs & subs without needing to transcode, I could possibly become interested. For now though - I'd just like to get Jellyfin working well and securely through Tailscale sharing. The TSDProxy solution looks fairly decent to me. I would need to read up a lot more on NPM and other solutions (Caddy is another I've heard of) before going down that route. The less open to the internet the machine is, the better. At least with TS, it is only visible to other TS users if I share them in. That means only a few devices in the entire globe can ever "find" my TS instance. That sounds fairly secure to me...

OK, this looks like a great option, so I decided to give it a try. I put Jellyfin container back into Bridge mode. I set up TSDProxy & Label Manager in unRAID. TSDProxy kept stopping. I think because I didn't have HTTPS enabled in Tailscale. Once I enabled that it seemed to stay running. Jellyfin TSDProxy machine appeared in Tailscale Machines. However when dropping my phone onto 5G, connecting to Tailscale, 2 things happen. 1 - I cannot connect to Jellyfin using the TS IP address using either http or https. I am unable to use Tailscale DNS on my phone as it knocks out my Private DNS settings for Secure DoT DNS. Jellyfin is accessible from the LAN. 2 - Although I can't connect to the Jellyfin server via the app or web interface on my phone, the Tailscale App allows me to Ping the Jellyfin "machine". This shows a relayed connection, when all other devices on the LAN now show Direct Connection. I tried changing the Labal Manager settings (Advanced) to https, and with 8096 in the container port, but this didn't seem to help either. TSDProxy logs show several errors, including 1:22AM ERR error host=x.x.ts.net method=GET module=proxymanager proxyname=jellyfin status=502 url=/ http: proxy error: read tcp 172.17.0.5:33764->172.17.0.1:8096: read: connection reset by peer Any ideas on what I'm missing?

Thanks for all the help. Yes, it's just Jellyfin I want to expose to the Tailnet. However, with my current setup (subnet available through router), every device on the LAN is potentially exposed to the Tailnet. That is fine for my usage, and how I want it for me as the admin. However I don't want that level of exposure to a shared device for family members. Not so much - more concerned if their personal devices were compromised or stolen. No point in having my entire LAN exposed. I tried reading up on the Tailnet ACLs, but was unsuccessful in creating a file that worked. I intend to return to this idea though. Hmm - that is also not something I am familiar with. Is it possible to create a DNS address like that, accessible only over my Tailnet? How would I link that to the Container IP address, and not the entire unRAID server?

OK, I have performed these actions. I enabled subnet LAN functionality on my router. I can achieve a Direct Connection to the router, and I assume it is a Direct Connection to other internal IPs, since it would be going through the router. My next question is about bolstering security. I would like to permit access to Jellyfin, via Tailscale, to other family users with whom I share the machine. However I do not want them having access to any other IP on the subnet. The "Use Tailscale" implementation made this easy - a Tailscale IP was supplied to the container, which I could provide to family. What would be the best way to implement this now? Or is that beyond the scope of this plugin?

Thanks for your replay @Espressomatic. I am using the Tailscale Plugin. I thought the sticky was referring to the Tailscale Docker, which I am not using. Apologies for posting in the wrong section. Unfortunately the sticky only mentions "Please make a post in the appropriate OS support forum". unRAID is my OS. What is the OS support forum I am requested to post in? I'll happily go there to ask my Jellyfin Docker integration questions. I thought exposing only the Jellyfin Docker to other users over the Tailscale integration was a secure method of sharing. I do not other users having access to any other part of the unRAID machine. Accessing the Jellyfin container is all those users will be permitted to do. Sure - I can access directly using the unRAID IP, but I will not be revealing that IP to other users. I am by no means a networking expert, just trying to make use of the functionality offered by the plugin. Which seems a lot simpler, and more secure, than allowing outside access by other means. At least for a less experienced user. But I'm happy to be pointed in a direction that makes better sense from a security standpoint, and that has easy-to-follow directions that would let me accomplish the same thing.

Quick question about this great plugin! I am able to connect directly when outside my network to devices inside my network, including unRAID. However I have a Jellyfin docker that I cannot reach directly. It always ends up in RELAY mode via its Tailscale IP. I can access unRAID through a Direct Connection. Guessing I messed something up with my config. Jellyfin docker has Tailscale Enabled. Tailscale shows on the Docker Advanced view, and the device is visible in Tailscale Machines. unRAID host is exposing its IP for subnet routing in the form x.x.x.x/32 - this has been approved in Tailscale web interface. Is there something I need to do (or change) to allow the Jellyfin docker to accept Direct Connections? Thanks

Files I forgot to attach in first post...

Having some issues after Docker.img ran out of space the other day. Temporarily increased space, but still having issues. Things starting to happen in the GUI. When using binhex-krusader docker, there are GUI issues. See screenshot in post below. When trying to view Docker logs, I just get a blank screen. When opening unRAID terminal, I just get a blank screen. Last night I rebooted the machine, it didn't come back online. Had to do a hard power off. Once it booted after the hard power cycle, it came back up in the usual amount of time. I can SSH into the machine fine. When running HTOP via SSH, I started seeing fatal errors listed. See screenshot. I don't know if all of this is related to Docker or not? Is the best thing to simply delete the Docker image and start again? I would also like to know how to figure out which Docker was writing to the docker.img and filling it up. I see my usage went up by another 100mb last night. Thanks for any hints / tips you can offer.

For now I have disabled chapter image extraction in Jellyfin settings, as I was seeing a lot of those. As for the Music folders, it looks like it was analyzing duration, not transcoding. At least I hope that to be the case. I would still like to drill down a little deeper though to discover the causes and find a good solution. All ideas welcome!

I recently switched over from binhex-jellyfin to bbergle-jellyfin, since I couldn't get hardware transcode to work on binhex version with the iGPU. Works fine in bbergle, but today I started receiving disk full warnings. Went from 95% to 99% over the course of a few hours. I disabled all dockers, turned on bbergle-jellyfin, and it started filling up again. A look at the logs from early morning shows : [2024-05-28 02:00:31.616 -06:00] [ERR] [17] MediaBrowser.MediaEncoding.Encoder.MediaEncoder: I-frame image extraction failed, will attempt standard way. Input: "file:\"/data/tvshows/HDTV/All Creatures Great and Small (2020)/Season 1/all.creatures.great.and.sma ll.2020.s01e01.1080p.bluray.x264-carved.mkv\"" MediaBrowser.Common.FfmpegException: ffmpeg image extraction failed for file:"/data/tvshows/HDTV/All Creatures Great and Small (20 20)/Season 1/all.creatures.great.and.small.2020.s01e01.1080p.bluray.x264-carved.mkv" at MediaBrowser.MediaEncoding.Encoder.MediaEncoder.ExtractImageInternal(String inputPath, String container, MediaStream videoSt ream, Nullable`1 imageStreamIndex, Nullable`1 threedFormat, Nullable`1 offset, Boolean useIFrame, Nullable`1 targetFormat, Cancell ationToken cancellationToken) at MediaBrowser.MediaEncoding.Encoder.MediaEncoder.ExtractImage(String inputFile, String container, MediaStream videoStream, Nu llable`1 imageStreamIndex, MediaSourceInfo mediaSource, Boolean isAudio, Nullable`1 threedFormat, Nullable`1 offset, Nullable`1 ta rgetFormat, CancellationToken cancellationToken) Similarly, after restarting the container after a few hours I found lots of logs like below : [2024-05-28 09:38:29.964 -06:00] [INF] [23] MediaBrowser.MediaEncoding.Encoder.MediaEncoder: Starting "/usr/lib/jellyfin-ffmpeg/ff probe" with args "-analyzeduration 200M -probesize 1G -i file:\"/data/movies/Music/The Ultimate Christmas Album/41 - The Cadillacs - Rudolph The Red-Nosed Reindeer.mp3\" -threads 0 -v warning -print_format json -show_streams -show_format" [2024-05-28 09:38:30.014 -06:00] [INF] [24] MediaBrowser.MediaEncoding.Encoder.MediaEncoder: Starting "/usr/lib/jellyfin-ffmpeg/ff probe" with args "-analyzeduration 200M -probesize 1G -i file:\"/data/movies/Music/The Ultimate Christmas Album/28 - Barry Gordon - Nuttin' For Christmas.mp3\" -threads 0 -v warning -print_format json -show_streams -show_format" [2024-05-28 09:38:30.067 -06:00] [INF] [33] MediaBrowser.MediaEncoding.Encoder.MediaEncoder: Starting "/usr/lib/jellyfin-ffmpeg/ff probe" with args "-analyzeduration 200M -probesize 1G -i file:\"/data/movies/Music/The Ultimate Christmas Album/38 - Four Seasons - Santa Claus Is Coming To Town.mp3\" -threads 0 -v warning -print_format json -show_streams -show_format" [2024-05-28 09:38:30.130 -06:00] [INF] [24] MediaBrowser.MediaEncoding.Encoder.MediaEncoder: Starting "/usr/lib/jellyfin-ffmpeg/ff probe" with args "-analyzeduration 200M -probesize 1G -i file:\"/data/movies/Music/The Ultimate Christmas Album/04 - Bobby Helms - Jingle Bell Rock.mp3\" -threads 0 -v warning -print_format json -show_streams -show_format" [2024-05-28 09:38:30.190 -06:00] [INF] [24] MediaBrowser.MediaEncoding.Encoder.MediaEncoder: Starting "/usr/lib/jellyfin-ffmpeg/ff probe" with args "-analyzeduration 200M -probesize 1G -i file:\"/data/movies/Music/The Ultimate Christmas Album/50 - Jimmy Beaumon t & The Skyliners - Another Lonely New Year's Eve.mp3\" -threads 0 -v warning -print_format json -show_streams -show_format" [2024-05-28 09:38:30.244 -06:00] [INF] [33] MediaBrowser.MediaEncoding.Encoder.MediaEncoder: Starting "/usr/lib/jellyfin-ffmpeg/ff probe" with args "-analyzeduration 200M -probesize 1G -i file:\"/data/movies/Music/The Ultimate Christmas Album/26 - Dion - Please Come Home For Christmas.mp3\" -threads 0 -v warning -print_format json -show_streams -show_format" Can anyone here advise me on how to troubleshoot further, or limit what is happening to fill up space, and how to free the space back up?

Well, I restarted Jellyfin server and Krusader. Disk usage from this morning to now has increased to 99%. It can only be one of these 2 culprits. I'm guessing Jellyfin. From the Jellyfin Docker logs I am seeing : [09:38:30] [INF] [33] MediaBrowser.MediaEncoding.Encoder.MediaEncoder: Starting /usr/lib/jellyfin-ffmpeg/ffprobe with args -analyzeduration 200M -probesize 1G -i file:"/data/movies/Music/The Ultimate Christmas Album/26 - Dion - Please Come Home For Christmas.mp3" -threads 0 -v warning -print_format json -show_streams -show_format [09:38:30] [INF] [33] MediaBrowser.MediaEncoding.Encoder.MediaEncoder: Starting /usr/lib/jellyfin-ffmpeg/ffprobe with args -analyzeduration 200M -probesize 1G -i file:"/data/movies/Music/The Ultimate Christmas Album/25 - The Beach Boys - Frosty The Snowman.mp3" -threads 0 -v warning -print_format json -show_streams -show_format [09:38:30] [INF] [24] MediaBrowser.MediaEncoding.Encoder.MediaEncoder: Starting /usr/lib/jellyfin-ffmpeg/ffprobe with args -analyzeduration 200M -probesize 1G -i file:"/data/movies/Music/The Ultimate Christmas Album/43 - Johnny Maestro & The Brooklyn Bridge - Christmas Serenade.mp3" -threads 0 -v warning -print_format json -show_streams -show_format [09:38:30] [INF] [33] MediaBrowser.MediaEncoding.Encoder.MediaEncoder: Starting /usr/lib/jellyfin-ffmpeg/ffprobe with args -analyzeduration 200M -probesize 1G -i file:"/data/movies/Music/The Ultimate Christmas Album/33 - Harry Simeone Chorale - Little Drummer Boy.mp3" -threads 0 -v warning -print_format json -show_streams -show_format [09:38:35] [INF] [27] Jellyfin.Plugin.KodiSyncQueue.EntryPoints.LibrarySyncNotification: Started library sync [09:38:38] [INF] [27] Jellyfin.Plugin.KodiSyncQueue.EntryPoints.LibrarySyncNotification: Library Sync: Updated 98 items Looks to me like it is reencoding all my music for some reason, and saving the new encodes in the container perhaps? Anyway, I'll refer to the specific container Docker support page and see what I can find out. I remembered I just switched Jellyfin dockers when I upgraded the hardware. This container behaves differently than the former it seems... Thanks for the tip on resizing the img file.

Rebuilt my machine not long ago, and added some new dockers. Namely Jellyfin, Immich, PostgreSQL_Immich. I woke up this morning to alerts from unRAID regarding docker.img dusk utilization. 3:18am it hit 93% 4:09am 94% 4:57am 95% 6:01am 96% I immediately stopped all Dockers. I have a 20GB .img file. According to Docker Volume Info : Total devices 1 FS bytes used 17.86GiB devid 1 size 20.00GiB used 19.52GiB path /dev/loop2 Total to scrub: 18.27GiB Rate: 0.00B/s Error summary: no errors found I've attached Container Size Report. I read over the Docker FAQ, but still have a few questions. My Jellyfin Media library is quite large. Lots of photos & videos for Immich as well. Jellyfin had been added about a month ago. Was previously using Kodi. Immich added a couple of weeks ago. Questions : How do I increase the size of my Docker.img file? How do I determine which Docker was performing write actions through the night? If it is log files, can I simply add the line to each container as per this Docker FAQ? If it is not log files, how can I determine what container is using up all the extra space? Thanks for your help follow unRAID fans!

I've got a transcoding question too. I see instructions for setting up nVidia HW transcoding in the Docs, but nothing for Intel QSV. I wonder if that means one of two things. 1 - Intel hardware transcoding does not work with this docker. 2 - Intel hardware transcoding should "just work" without additional configuration. From what I'm reading here, multiple people are having problems with Intel QSV transcoding. Can anyone confirm if they've got it working? Or have people moved over to the Bbergle-jellyfin container instead?

Thanks for the reply @JorgeB. It is a btrfs SSD. I was worried the appdata, system data etc might not remain fully intact if I stop the array, unassign the SSD and follow the rest of the procedure. However I ran the Appdata Backup plugin, invoked the mover (so backups ended up on the array), and then followed your instructions. Everything worked perfectly. Now I can view my cache utilization at a glance again 🙂 Thank you for your help!

I was swapping a cache disk - but using the pool method to do it. Now I need some help for the final piece of the puzzle! Initial cache disk - old 2TB Seagate Barracuda. New cache disk - old 500GB SSD. Shutdown Docker service. Added SSD to unRAID, and added it to the cache pool. Cache pool assigned SSD the name cache 2, made a pool, and balanced. I unassigned the old spinner cache disk, and the pool rebalanced again, as expected. Shut down unRAID, disconnected the SATA power cable from the 2TB spinner, rebooted. Restarted Docker, tested containers, everything good. When I look under MAIN in the menu, scroll to cache, it lists Pool Devices. The old 2TB HDD is still kind of showing. It says cache (the original name), not installed. Then beneath that, cache 2 with the SSD drive. Almost everything works, /mnt/cache is full of the data it is supposed to have, but I would like to completely remove the old cache entry for the 2TB drive that is now gone, so the Pool only shows 1 device. I am hoping by doing so, I will regain a Utilization graph on the Dashboard page of the cache. Right now, below Array, it says Cache and tells me the correct figure. However the Utilization bar remains empty. So it seems I got 90% of the way through, but perhaps need to do something else to finally remove the old drive from the cache pool. Have I missed a step, or what is it I need to do? And by the way, is it the Appdata Backup by Robin Kluth that is the recommended way to backup the Appdata? It seemed older posts were referring to another App which is no longer in the CA app store.

Well this is embarrassing... replacing the inadvertent comma fixed me up right away. How did I not see that? Thanks for spotting it @ConnerVT !

You think it is ZT blocking it? I can access port 6080 for binhex-krusader & 8384 for syncthing on the same IP address. So docker ports seem to be working in general. Just not delugevpn.

Lol - yes the x is a 0 - I obfuscated the wrong section At any rate, with it being a zero, I think it is correct. It is working fine when I am on the physical network, so I know that bit is alright. Am I still missing something in having added the ZeroTier address range? 192.168.195.0/24

Problem using with ZeroTier : I've been using binhex-delugevpn for years. Works great on my LAN. Right now I am working away from home, and using ZeroTier to connect to my unRAID server. I can view the regular unRAID web interface with no problems. Unfortunately I am unable to view the DelugeVPN WebUI, or connect to it via the Deluge client on my laptop (using the same IP address as the unRAID server, as if I was on the same LAN (through ZeroTier). I have added the ZeroTier IP range to LAN Network key, which now reads, 10.0.10.x/24,192,168.x.0/24 (where x is a number removed from this post for security reasons). Any thoughts on what I might be doing wrong here?

Thanks Jonathan. This all worked without error. I was able to create and see the test files (I actually created multiple, in the incomlete / complete folders Deluge uses). In the end I decided to simply erase all files and delete the share. I then created a new share, pointed Deluge in the right direction, and everything is functioning again. Not sure why I had to delete and recreate a share, but it is now working again :-)

Still having no luck with this. I deleted the Docker container, re-installed, and am still experiencing the same issue. Torrents show up, it looks like it tries to download them, but download speed never shows anything, despite seeing an extensive list of seeds & peers. I'm wondering if it is some kind of disk access issue. The network side appears fine. However when I start a download, no files ever appear in the /data folder (mapped to mnt/user/download/incomplete). Could that be the issue, and if so how would I troubleshoot?