No index.html results in 403

[reavessm]

Command run:

docker run -d \
-p 443:443 \
-v /local/path:/config \
-e PUID=XXXX \
-e PGID=XXXX \
-e TZ=America/NewYork \
linuxserver/codiad"

When I log into 'https://localhost', I am greeted with the nginx '403 Forbidden' page. After digging through the code I found that /config/www/index.{html,php} does not exist. I then ran echo "Hello" >> /local/path/ww/index.html and revisited the site. Then I could see the 'Hello' message. This also seemed to happen when using HTTP as well as HTTPS.

I was running this on Gentoo Linux, docker version was 18.09.5, image version was latest.

[tobbenb]

No point in anonymising you PUID and PGID. It's also good for us to have fusing troubleshooting.

1. You have mapped the wrong port. It's set to use port 80.
2. You haven't set the host path for the /config volume mount correctly.
3. You have a " at the end. Remove it.
4. You forgot to set the name. --name=codiad

[reavessm]

1. Your dockerfile exposes both ports 80 and 443, but I am getting the same result when mapping port 80 while using HTTP and mapping port 443 while using HTTPS
2. Replace "/local/path" with "/mnt/VMStorage/Codiad" in my OP if you want my actual setup. I also set the PUID and PGID to be 1001. "/mnt/VMStorage/Codiad" is owned by 1001:1001 and has permissions of 775. I can wipe the directory, recreate the container, and it gets repopulated, so the container has write permissions
3. Typo, I didn't actually run it with that command
4. I did set that, just forgot to add it to the issue because it shouldn't affect anything, right?

I have tried to access this via an existing reverse-proxy, as well as locally. Both have the same result.

I can get the container running, the issue, at least I think, is because the '/config/www' directory only has an empty 'plugins' directory and nothing else.

[j0nnymoe]

Assuming /mnt/VMStorage/Codiad is a remote mount, Try putting it on your local file system.

[reavessm]

@j0nnymoe That worked! I created a ~/Codiad folder, mapped that to config, and it came up on the first try! But I have other containers running on that NFS share just fine? My ~/Codiad directory actually has less permissions (755 vs 775 with the other).

[j0nnymoe]

Yea we get this alot. We don't recommend putting /config mounts because stuff that uses sqlite etc don't like it.

[reavessm]

@j0nnymoe is there a sub directory that I can mount on NFS shares to hold the source files? Can I map like /config/projects on the share? I'm fine with user config stuff being wherever, but if I have a lot of projects, I don't want them cluttering things up

[reavessm]

Actually, this might not have been fixed. It's been trying to create the initial project for about 10 minutes now, and it's still spinning. Also the /config/projects directory is completely empty

[j0nnymoe]

Only thing I can suggest is try it and see what happens. Add in -v /path/to/host:/config/projects.

[CHBMB]

Just a heads-up @reavessm

We're in the process of deprecating this and releasing an alternative IDE as Codiad is no longer maintained upstream.