bdillahu

Just to report in - worked for me... I had my networking on the old container set to br0 and a fixed IP - it wouldn't adopt in the new world until I made that match (might want to mention that in the instructions) - otherwise, seems straightforward. Thanks for doing this - didn't really want to run two containers 🙂

I've been using the new tailscale ssh feature (works great), and ran into something with the docker on unRaid that I don't quite understand. If I start tailscale up with the ssh flag, it starts. But when I connect to the IP for my unRaid server (where the docker is running), I am put in a session "inside" the docker, not inside the unRaid OS. I kind of see why, but I'm not sure how to change things so that I get more of what I'm expected. Anybody have a suggestion? Thanks!

Same thing with a photo app (Mylio) - it's the initial "setup" screens, which in this case are a "popup" kind of graphics thing... I tried using VNC and Mac screen sharing - same issue. Edit: I also tested with Chrome and see the same as you.

Same scenario... flipped from VNC to a passthrough (which didn't work) and trying to get back to working VNC. Had to rebuild one. Another is a Hackintosh that I'm still trying to repair.

I use offlineimap to download all my accounts - running on a linux VM, haven't done it in a docker.

In the docker section, I notice that if I set the "all" switch to off so that I can pick and choose, my last Docker doesn't appear... (It's Wallabag FYI). not sure if it's specific to that docker, or an "off by one" kind of error.

Glad you got it going!

I don’t think so, but not sure. I think that’s the userid inside the container, and the PUID stuff is what is presented to the host file system

Note, I just editing my find command, sorry, swapped the parameters

Assuming normal unRaid setup... I only know how to do this via command line - there might be some other way, but I don't know. log into the box > ssh <your machine> Assuming your config above (and that syncthing is empty): > chmod 777 /mnt/user/syncthing > chown nobody:users /mnt/user/syncthing If there are files in /mnt/user/syncthing already: > find /mnt/user/syncthing -type d -exec chmod 777 {} \; > find /mnt/user/syncthing -type f -exec chmod 666 {} \; > chown -R nobody:users /mnt/user/syncthing Obviously, no guarantees 🙂

Hmmm... I don't know, sorry... your config looks about like mine, and mine has been working fine. I have fits with permissions on unRaid for some reason - I wish it did things a bit more "standard" (at least as I understand standard), but... Have you tried changing the permissions of your target directory? Make sure all directories are 777 and files are 666, including (in your config) /mnt/user/syncthing itself? And the ownership should match the 99/100 of PUID/PGID. Otherwise, sorry, all I know to check.

Your PUID value is wrong... you have put the file permissions mask in, not the PUID - should be 99 I believe, on a normal unraid setup.

Hey, thanks for the idea... I went back through things and sure enough... If you actually read the help (:-)) "Local server uses NAT" "When NAT is enabled, the server uses its own LAN address when forwarding traffic from the tunnel to other devices in the LAN network. Use this setting when no router modifications are desired, but this approach doesn't work with Docker containers using custom IP addressess. When NAT is disabled, the server uses the WireGuard tunnel address when forwarding traffic. In this case it is required that the default gateway (router) has a static route configured to refer tunnel address back to the server." I had it turned on... turned it off, configured the recommended static route on my router and things seem happy.

I'm sorry to drag this back up... I left it for a while, thinking somebody with more skills than me might encounter it and maybe fix it, but apparently I'm "special" 🙂 When I have wireguard enabled, then I can't access my Unifi Controller docker (on br0) - don't have to be connected via Wireguard - it just has to be enabled. I turn it off and everything works great. Wireguard works fine when enabled, except for killing this one connection. I don't know how to debug it further... any thoughts? Appreciate it.

I tried changing appdata to cache - no change. I was using br0 with a static IP... changed to Bridge mode and can get to it on http, not https... scratch that, I lied... I can get on https... as long as I'm on bridge Seems to still be something with network side, I think.