C3004

👍Ok, I'll add it to the template to prevent this issue in the future.

Hi, at the bottom of the screen click on Add another Path, Port ... Configure it for your config, like this: Now press add and lastly apply and wait for the container to start. This should work. In the logs we see that it can't write to /comics/... which is the filepath the komga container sees internally. By adding it to komf it will have the access it did not have before. Can you tell me what komf settings you use? I use api mode so i didn't run into this problem, but i think you may be using the comic_info mode to update the metadata - which would need the extra access. Good Luck!, C3004

Hi magnuspraeda, This is the first time I ran into this Problem. For your issue i found this thread on Github: https://github.com/Snd-R/komf/issues/92 Maybe Kohegas Solution could work for you: Add your Library the same way to komf as it is in Komga. If that doesn't work please share your docker run for komf + komga. Have a nice Day, C3004

Now that Komf works for me again, i will make a new writeup or Video for the Komf setup (using the userscript).

Hi Ollebro, I ran into maybe the same Issue. The Userscript is configured to try to connect to Komf on localhost and 'forgets' your settings when you restart the script. So i recommend changing the adress in the Script itself (note that after updating your Script you will have to set this again) In Tampermonkey use Find and search for localhost. use find next next to the search bar until you see this: In Line 19787 replace http://localhost... with your Komf server adress. (192.168.1.4:8085) You can skip the http:// it worked without in my case. That should keep the Script from forgetting your komf adress (until the next script update) Also make sure you've set your komga and kavita adresses to let the script be active there. Good Luck, Have Fun

Please post your Docker Run. Note that i also need to see which repository you're running.

Weird. I don't see anything wrong. Do the container Logs currently note any errors? If the container is working correctly set the debug variable to false that the logs don't fill up the docker.img. Just for fun you could also run the docker system df command(https://docs.docker.com/engine/reference/commandline/system_df/) But the debug variable probably filled up the logs.

Hi Soldius, That variable is new and currently only in the preview containers. Please note that the fix only works with the develop container. I wrote about the Fix before the tachidesk guys made a full rewrite of the docker container. The rewrite happened two weeks ago. The Variable you want to use is part of that new rewrite. As the fix only works in the develop container and the latest update of that one was 18 Days ago the Variable is currently not available in that one. Best Regards

New Way to get Rid of the File permission error. I Don't know if this will stick, or if it will get added in the default container. Create a Backup before adding or removing these Changes. Don't use this if you made the webui publicly available. They added a function to the development container to change the running user. (https://github.com/Suwayomi/docker-tachidesk/issues/22) Activate advanced view to make these changes Change **Repository** from "ghcr.io/suwayomi/tachidesk" to "ghcr.io/suwayomi/tachidesk:develop" Add in **Extra Parameters ** "-u 99:100" without quotations It should look like this afterwards:

Hi Steven, i took a short look. Could replicate the error. Currently it seems to me that the docker won't accept that the Kavita fields are not filled in. The Docker normaly should ignore kavita if the event listener is not activated. I'll take a closer look later. Anyway the latest version without kavita would be this one sndxr/komf:0.12.0 but it's not kompatible with the new config.

Yeah please share your docker run config. If it gets replaced with an update it usually means that the files don't get exported from inside the container to your harddisk. But in a different case did you use Parag2K3's sugesstion for the permission error? If so the new user doesn't have acces to the config. But it seems like they already made a change that would fix it. See here: https://github.com/Suwayomi/docker-tachidesk/issues/22 I didn't test it but if it gets through it should be fixed in the next update. Have a nice Day

To be moved to DockerContainer support. Ask here for any Problems that might show up during Komf-Docker installation or usage. Komf replaces all Metadata from what you're modifying! Installation: It's the same like you know for other Containers. Fill in The required Fields and ignore the others when not needed. Config: In the current docker the application yml is not provided. So I put it here application.yml(31.12.22) for you . You need to put it in Komf's Appdata folder. There you can take a look at the advanced settings. Info: Komf is currently an API project only. Unless you know how to deal with those you can join me in using the Userscript. The Script adds a new button on the right of the usual config buttons in Komga: It will let you Identify the Series and add Metadata to it. Preparation Userscript: You will need a Userscript manager. I took Snd-R's recommended one named Tampermonkey. Download the Userscript from his Release-Section. Open the Settings from Tampermonkey. Drag and drop the js file in there. Press the install button. Now you should see the Script in the installed Section. Press the edit button on the far right. 1. Change the first URL to your Komga-Instance. 2. Change the second URL to your Komf-Instance. Now save the changes under File. You can now reload your Komga Tab and enjoy the Metadata Grabber. General Links: Github-Repro-Komf-Docker: https://github.com/Snd-R/komf Github-Repro-Komf-Userscript: https://github.com/Snd-R/komf-userscript Github-XML-Template: https://github.com/C3004/Unraid-Templates-C3004/blob/main/Komf-Docker.xml Github-Repro-XML-Templates: https://github.com/C3004/Unraid-Templates-C3004

Password Protection over Swag: In the swag docs they have this: If you'd like to password protect your sites, you can use htpasswd. Run the following command on your host to generate the htpasswd file: docker exec -it swag htpasswd -c /config/nginx/.htpasswd <username> You can add multiple user:pass to .htpasswd. For the first user, use the above command, for others, use the above command without the -c flag, as it will force deletion of the existing .htpasswd and creation of a new one source:https://docs.linuxserver.io/images/docker-swag (scroll down to Security and Password Protection) For that to work you will also have to uncomment these two lines in the config file. #auth_basic "Restricted"; #auth_basic_user_file /config/nginx/.htpasswd; Password Protection in the Docker: Just saw this in their discord server: You can also add simple authentication in the docker config. But i do not know how it will hold up in reference to the one directly in the reverse proxy.

I didn't think that Dockers work that way but apparently they do. For volume mapping the unraid host uses the -v command which adds the problem discussed bellow: According to the Dockerfile in the Tachidesk-docker repository(https://github.com/Suwayomi/docker-tachidesk/blob/main/Dockerfile). Their Suwayomi user has UID and GID set to 1000. My thoughts with dockers were that the id in the docker does not matter so the id's of the user running docker would be used. But this does not happen. This is my ls-l output from the download folder: root@Tower:/mnt/user/Mediaserver/Manga/Tachidesk/MangaDex (EN)/This Village Sim NPC Could Only Be Human# ls -ltotal 0 drwxr-xr-x 1 windows 1000 224 Jan 17 13:22 Kirei\ Cake_Ch.16.2/ drwxr-xr-x 1 windows 1000 306 Jan 13 17:20 Kirei\ Cake_Ch.17.1/ drwxr-xr-x 1 windows 1000 306 Jan 13 15:12 Kirei\ Cake_Ch.17.2/ drwxr-xr-x 1 windows 1000 276 Jan 13 15:11 Kirei\ Cake_Ch.18.1/ drwxrwxrwx 1 nobody users 276 Nov 7 20:42 Kirei\ Cake_Ch.18.2/ If you look at it you see two different users. The Nobody user is there from when i tested the chwon and chmod command at the system. But in the other files we can see that a different user created them. It is the windows user i created in Unraid to access the drives. Now let us see what output less /etc/passwd gives me: windows:x:1000:100:normal windows login:/:/bin/false Now if we remember to the beginning the suwayomi user has the UID set to 1000, which is the same as my windows user. So the docker was only working on my System because some USERID was the same as the USERID from the Docker. ---------- How to fix it. OPTION 1:(Prefered) I contact the Tachidesk guys and have them add some option we can use to change the ID's the docker uses to make/change stuff. OPTION 2: If Option 1 fails I'll make a fork and set it up to use changes in option1 and that it updates automatically from the suwayomi repository. ------- Ok it might take a bit till this is sorted out fully. I have exams and a paper comming up so it could be that it is fixed at the end of February. Till the you can use the docker without any added paths. Important: When the docker has an update available do not update it immediately! Create a Backup file in the Settings and save it. Then you can update and use the backup file to get your progress back.

Ok, looks like I did not do my Research correctly. I added the PUID and GUID variables to the container template as i thought that those would be handled by the docker engine but apparently the docker itself needs to have support for it so it doesn't work in our case. They are currently useless and i will remove them from the Template. Ok so the Problem is that the docker can't write it's files. A few questions i have for you: What are the zpool share settings? Is It shared with smb? (The permissions can get funky if windows has access to it.) Is the zpool/tank_docker/appdata path set as standard in the docker settings?(If not maybe the docker engine gets no access to it) things to try out: -Removing the docker from the unraid system and reinstalling it. -Removing all the mount paths to see if the docker runs like that. (With removed filepaths the docker will only run in it's own container and does not keep data when updating the container) If it does not work then something different is wrong on your end. That are my thoughts for now. I'll do some more research later.