April 6, 20251 yr Author ok, I manage to have a server diagnostic. servraid-diagnostics-20250406-0950.zip
April 6, 20251 yr Community Expert As mentioned, you should uninstall Mover Tuning. The version you have installed has been deprecated anyway. Also these ca.mover.tuning.plg - 2023.12.19 (Deprecated) (Up to date) disklocation-master.plg - 2025.02.08 (Unknown to Community Applications) folder.view.plg - 2024.10.02 (Update available: 2025.01.11.1) (Incompatible)
April 6, 20251 yr Community Expert 20 hours ago, Nodiaque said: System is the one that's about 650gb by itself system shareUseCache="prefer" # Share exists on cacheappdata Probably you want appdata and some of those others mentioned moved to that same pool. But first, I would like to try to understand why system share is so large. Normally, that user share only has docker.img (yours is 60G which might be reasonable) and libvirt.img, which is default 1G and never any reason for that to be larger. So, what do you get from command line with this? ls -lah /mnt/cacheappdata/system
April 6, 20251 yr Author folder view is the version compatible with my versino of unraid. That's why I can't update. It was update when I upgraded the server to my current version. But new version got out for 7.x and it started showing incompatible for the new version. disklocation-master is just a simple app that give me a little gui of where the disk are physically installed. Maybe it was shutdown with 7.x It's still active, even have an update in march located here: But I do see it cannot check for update right now. Might just have to install new version when I'll upgrade to 7.x. All of these are to be checked when I'll jump to 7.x. But not what is causing the underlying issue though. You might see a log of no more space or something. I ran mover during the night and /var/log filled up because of the logging which caused issue this morning. I disabled mover logging, stop/start the array to clear var/log and restarted mover so it move stuff back to cache to be able to run docker Edited April 6, 20251 yr by Nodiaque
April 6, 20251 yr Author Just now, trurl said: system shareUseCache="prefer" # Share exists on cacheappdata Probably you want appdata and some of those others mentioned moved to that same pool. But first, I would like to try to understand why system share is so large. Normally, that user share only has docker.img (yours is 60G which might be reasonable) and libvirt.img, which is default 1G and never any reason for that to be larger. So, what do you get from command line with this? ls -lah /mnt/cacheappdata/system Yes they will all move, now the problem is the system too large. My docker is set to use folder not img.
April 6, 20251 yr Community Expert 1 minute ago, Nodiaque said: My docker is set to use folder not img Did you do this because docker.img kept growing? docker.img (or docker folder) is for the executable code of the containers, and shouldn't really be growing except for a small amount when you install a new container. The usual reason this grows is because one or more containers are writing to paths that aren't mapped to host paths.
April 6, 20251 yr Author I had an issue that I really can't remember and I was told to switch to folder and it solved the problem. Before that, I think I redid more then once the img which caused me to lose everything each time edit: I do have an issue right now where I was asked to switch to img back. I have many docker that when I do stop in the webgui, it says error after the timeout but it still shutdown the docker. Using docker stop shutdown the docker nearly instant. But that's another topic. Edited April 6, 20251 yr by Nodiaque
April 6, 20251 yr Community Expert Just now, Nodiaque said: switch to folder view and it solved the problem. Before that, I think I redid more then once the img which caused me to lose everything each time switching to folder view probably didn't really solve the problem, it just hid it. No reason at all to lose everything when you recreate docker.img. Previous Apps will put them back just as they were. What do you get from command line with this? du -h -d 1 /mnt/cacheappdata/system
April 6, 20251 yr Author ouf, sit back, take a beer, coffee or whatever drink float your boat. This command will take time. Running it
April 6, 20251 yr Community Expert I don't know what remoteviewer is but it isn't taking significant space. As you can see, your docker folder is the main problem. Enable Docker in Settings. Then go to Docker page, click CONTAINER SIZE button at bottom, and post the result.
April 6, 20251 yr Author Weirdly is that when I did the move, the cache drive had 380gb used space. I move it to the new folder and bam, it's now 700gb all the while docker was shutdown. Wild. I currently have mover running to put back appdata on ssd pool. I'm starting docker afterward to see if it still start or not.
April 6, 20251 yr Community Expert https://docs.unraid.net/unraid-os/troubleshooting/docker/#docker-image-full 34 minutes ago, trurl said: No reason at all to lose everything when you recreate docker.img. Previous Apps will put them back just as they were. https://docs.unraid.net/unraid-os/manual/docker-management/#re-create-the-docker-image-file https://docs.unraid.net/unraid-os/manual/docker-management/#re-installing-docker-applications https://docs.unraid.net/unraid-os/manual/docker-management/#docker-custom-networks
April 6, 20251 yr Community Expert That time consuming process is probably what I am going to suggest to get your docker.img under control. Maybe if we can figure out which containers are filling it up you won't need to do it again. I have about 20 containers and they easily fit in 20G docker.img. No way even 100 containers should take over 600G
April 6, 20251 yr Author Nah I know it's not normal, it wasn't taking anything near that before I move them to the new pool while offline. I never boot them back since I did the move. Doing it right now to see at least if they work. Edited April 6, 20251 yr by Nodiaque
April 6, 20251 yr Author ok, so everything is working fine. Although I don't like the 70C on my nvme drive though. Edited April 6, 20251 yr by Nodiaque
April 6, 20251 yr Community Expert 2 hours ago, trurl said: Enable Docker in Settings. Then go to Docker page, click CONTAINER SIZE button at bottom, and post the result.
April 6, 20251 yr Author I wanted to upgrade some share cause some point to cache directly, and I get a btrfs snapshot error Name Container Writable Log CodeProject.AI_ServerGPU 25.6 GB 14.1 GB 8.31 MB CloudflareBypassForScraping 5.71 GB 1.62 GB 6.60 kB immich 5.62 GB 80.4 MB 496 kB homeassistant 3.43 GB 1.60 GB 23.8 kB sonarr 3.13 GB 1.23 GB 1.04 MB calibre 2.94 GB 34.4 kB 23.7 kB FileFlows 2.69 GB 786 MB 134 kB duplicati 2.44 GB 1.97 GB 82.8 kB ConanExiles 2.37 GB 3.16 MB 4.63 MB V-Rising 2.35 GB 0 B 120 MB calibre-web-automated 2.21 GB 333 MB 819 kB Enshrouded 2.18 GB 2.69 MB 5.20 MB epicgames-freegames 1.54 GB 30.6 MB 427 kB nextcloud 1.39 GB 267 MB 20.5 kB Jellyfin 1.27 GB 22.1 MB 6.21 MB LyrionMusicServer 1.11 GB 0 B 120 MB free-games-claimer 1.03 GB 0 B 120 MB spotizerr 1.02 GB 2.69 kB 2.29 kB emulatorjs 866 MB 0 B 120 MB calibre-web-automated-book-downloader 863 MB 37.2 MB 3.52 MB calibre-web-automated-book-downloader-fr 863 MB 37.1 MB 3.52 MB LMS 859 MB 339 MB 7.12 kB youtube-dl-material 806 MB 115 MB 51.4 kB openHAB 799 MB 124 MB 496 kB flaresolverr 709 MB 162 kB 2.29 MB postgresql_immich 692 MB 63 B 31.4 kB Emby 679 MB 4.54 MB 1.05 MB All-The-Mods-9 675 MB 0 B 466 kB grafana 663 MB 0 B 858 kB swag 658 MB 254 MB 105 kB code-server 629 MB 25.1 kB 56.6 kB NodeRed 601 MB 0 B 1.66 MB unifi-network-application 591 MB 58.2 kB 46.2 kB romm 562 MB 2.48 kB 2.81 MB telegraf 551 MB 30.4 MB 147 kB cups 530 MB 12.6 kB 1.14 kB hydroqc2mqtt 495 MB 0 B 8.16 MB GameVault 470 MB 2.27 MB 8.82 MB mariadb-nextcloud 465 MB 0 B 50.9 kB UptimeKuma 440 MB 217 kB 860 kB postgresql15-gamevault 429 MB 63 B 926 kB unifi-bd 427 MB 0 B 9.10 MB Influxdb 406 MB 0 B 4.07 MB bazarr 396 MB 76.3 kB 8.88 MB GameManBot 366 MB 0 B 1.17 MB ombi 340 MB 43.0 kB 1.60 MB mariadb 336 MB 2 B 325 kB RedisInsight 304 MB 0 B 282 B jackett 292 MB 113 MB 339 kB MQTTExplorer 282 MB 0 B 26.5 kB MKVToolNix 268 MB 20.9 kB 180 kB MediaInfo 213 MB 42.6 kB 1.06 MB radarr 207 MB 22.6 kB 8.50 MB qbittorrent 196 MB 23.1 kB 13.2 kB QDirStat 184 MB 33.2 kB 495 kB prowlarr 183 MB 22.7 kB 1.12 MB Palworld 164 MB 42.6 MB 42.1 kB cross-seed 164 MB 0 B 3.40 MB Speedtest 148 MB 122 B 988 kB redis-nextcloud 145 MB 109 kB 120 kB redis 144 MB 109 kB 344 kB RustDeskServer-AiO 143 MB 4.37 kB 430 kB pihole 138 MB 48.2 MB 147 kB OpenTTD 126 MB 0 B 120 MB adminer 121 MB 23.1 kB 6.85 kB jellyplex-watched 118 MB 15.1 MB 835 kB netbootxyz 118 MB 23.4 kB 40.7 kB evcc 102 MB 0 B 4.14 MB qbit_manage 97.2 MB 271 kB 818 kB authelia 72.4 MB 199 B 698 kB dockersocket 19.3 MB 4.10 kB 142 kB unpackerr 16.2 MB 0 B 4.32 MB mosquitto 9.90 MB 57 B 4.56 kB nebula-sync 6.13 MB 0 B 234 kB Total size 89.2 GB 23.2 GB 717 MB Edited April 6, 20251 yr by Nodiaque
April 6, 20251 yr Author I'm going to clear my docker folder. I think it's the best thing to just recreate it.
April 6, 20251 yr Community Expert Why do you have both LMS and LyrionMusicServer? Or is LMS something else? I used LMS for years but switched recently and after a little time figuring out Lyrion I got rid of LMS. It only serves music to my Squeezebox Radio Alarm clock. Your sonarr seems too large. On my system, radarr and sonarr are similar in size. You have many I haven't tried, and some others I have but in the end decided I didn't need them.
April 6, 20251 yr Author LYrion is offline. I tried lyrion to replace lms and it wasn't working at all. So I just kept it to update later and try switching back. Only LMS is used for now. My sonarr is handling over 3000 tv shows while my radarr only 1000, that might be why. edit: I think radarr and sonarr here are not from the same person, might have different base. I'll check once I restore everything. Still waiting on the delete to finish Edited April 6, 20251 yr by Nodiaque
April 6, 20251 yr Community Expert Just now, Nodiaque said: My sonarr is handling over 3000 tv shows while my radarr only 1000, that might be why. I'm about the opposite, movies vs tv.
April 6, 20251 yr Author cannot say. Maybe database bigger because of the number of files (tv shows have way more files then movies). If you think put a medium of 10 episodes per series and 1 season each, it give 30 000 video file (plus all the extra file). But since many have way more then 10 episodes and more then 1 seasons, this number is shy from the truth. Although now that I think of it, the database should be in the appdata like the log file. I don't know. Maybe indexer temp. Anyway, it's deleting right now.
April 6, 20251 yr Community Expert 11 minutes ago, Nodiaque said: the database should be in the appdata This is generally true for all containers. Working data in appdata, with additional mappings to access other storage on the host for reading and writing. docker.img (or folder) should mostly just be the downloaded installation of the code for the application. Maybe some dockers don't behave nicely and do things in the recommended way. Maybe some have working data scattered all over the internal OS. But the troubleshooting link I posted above has things to pay special attention to when specifying paths in the docker template (installation) and paths within the application. Within an application, if a path is specified that is not in a container volume which is mapped to a host volume, then that path is inside docker.img. Container and Host volume paths must begin with /, and are case-sensitive.
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.