MarkusMcNugen

You shouldn't need to create the users manually. You can just them in the config file and the container will create them automatically. If the SSH key exists for the user it will copy it to their authorized_keys file as well. I'm glad you got it figured out at least. Sorry I didn't have team to look I to it further this week. Was waiting for the weekend to test it myself.

Are you mounting that folder to the docker host? If not then it wouldn't have persistence. There is already a volume for persistence that you can use for the container. Check out the docker hub or within pages for the container. You should have /config mounted to the docker host and your ssh key for the SFTP server should go in /config/sshd/keys. You'll probably need to be logged in as root on your Unraid server to get into that folder once you have /config mounted, its locked down pretty hard for security.

SFTP users are jailed in their home directories. If you want to share that folder with him you have to mount /mnt/users/videos inside the home directory for his user profile in the container, something like /home/*usersname*/videos. Here's an example with a user named josh and a music folder: https://github.com/MarkusMcNugen/docker-sftp#sharing-a-directory-from-your-computer I'm sure you can see how this would get rather cumbersome the more users and folders you have. Which is why you can also use a bash script to mount directories in users home folders that are mounted somewhere else inside the container. It's all on the github and dockerhub readmes. If you have any specific questions, ask away. https://github.com/MarkusMcNugen/docker-sftp#bindmount-dirs-from-another-location The thing is, since you have access to the sshd_config file you can make edits to it with the unraid root user where you have the sftp container mounted to your host, so you can configure it however you want. Dont want users jailed to their home directory? Want to jail them in a different directory? Want to filter based on user group and apply some options to specific users? All up to you

Well that sucks, sorry I couldnt help more. If you are just looking for a cert for HTTPS than just use the Swag container with LetsEncrypt and SNI and forward with the reverse proxy to the crushftp http interface. You only really need the LetsEncrypt plugin with CrushFTP if you plan on doing direct HTTPS access from the outside or FTPS. I personally dont use FTPS, I just stick with SFTP and call it good. It's all about the SSH keys at that point instead of a SSL cert.

Only other thing I can think of would be to check the CrushFTP.log file and see if it lists some kind of error when trying to read the plugin. I use letsencrypt for HTTPS with the swag container but purchase certs for other things. You can get some super cheap certs these days, ssls.com is where I buy mine. https://www.ssls.com/ssl-certificates/comodo-positivessl

Did you use the correct plugin file, maybe you grabbed the CrushFTP 9 version? I copied the CrushFTP 10 version to the plugin directory, restarted the container and it's there. 😉 CrushFTP LetsEncrypt Plugin: https://www.crushftp.com/crush10wiki/attach/LetsEncrypt plugin/LetsEncrypt.jar

Let me do some testing and see if I can get it to work myself. I can tell you that the base alpine container it's running on uses Java 16. My setup is a bit funky since I dont want to use a non-standard HTTPS port and I already have a bunch of existing sites behind a reverse proxy. So I use the Swag container with lets encrypt to reverse proxy external HTTPS with SNI to CrushFTP for webdav access and have the other ports forwarded directly to the container through my firewall. That doesnt match up with everyone's use case so Ill need to do some testing myself to see what issues there may be with that CrushFTP plugin and Ill let you know.

If it's working with chrome but not with edge then that leads me to believe it's a cookie or cache issue with the edge browser. If you havent already try clearing those and see if it's working. BTW, you are all good. You dont have to say what container you are talking about, I was asking fasur87 because it wasn't entirely clear and they didn't provide much info lol.

Server is starting just fine. What is the network mode set to? Host? Bridge? br0? However you are accessing the server it's sending a ridiculous HTTP Header that looks like it's meant for Unraid.

Which container? SFTP? Would help if you could post the logs from the running container.

That's expected output. Can you send, attach, or post the CrushFTP.log file from the appdata folder?

No changelog. When I update a container that's simply what it is unless something is broken. I try to design containers so nothing has to manually be done to build an update other than run the build on Dockerhub. I've started using test tags and branches to test updates now after the whole SFTP debacle. For example, when a build is initiated for qbittorrentvpn it dynamically grabs the latest version number of qBittorrent from their website and then downloads and installs that version. This way nobody has to manually edit the Dockerfile and change anything.

Hey Squid, The container has been fixed, upgraded, and put back up on dockerhub in perfect working order. Is there anything I need to do to get it unblacklisted? Thanks!

Not you, the container was blacklisted because it was broken. The base image it's built on actually broke syslog-ng because of an error they made in it's config file. I didnt have any time to put toward fixing it at the time. I actually just pushed an update that fixes everything and includes a few small improvements.

I will see if I can do this over the weekend. I usually just build from dockerhub with auto triggers and call it good but I have a VM I use for building as well.