# Docker (All platforms)

A Docker image exists for deploying LANraragi installs to your machine easily without disrupting your already-existing web server setup.\
Docker is the best way to install the software on remote servers. I don't recommand it for Desktop machines and casual users due to it being a bit complex to wield.

## Cloning the base LRR image

Download [the Docker setup](https://www.docker.com/products/docker) and install it.

{% hint style="warning" %}
The LRR Docker container uses a fairly recent ([3.14](https://alpinelinux.org/posts/Alpine-3.14.0-released.html)) version of Alpine Linux as its base. I recommend you use at least Docker version **20.10.0** to avoid issues with the `faccessat2` syscall.\
You can check your Docker version by executing `docker version`.
{% endhint %}

Once you're done, execute:

```bash
docker run --name=lanraragi -p 3000:3000 \
--mount type=bind,source=[YOUR_CONTENT_DIRECTORY],target=/home/koyomi/lanraragi/content \
--mount type=bind,source=[YOUR_DATABASE_DIRECTORY],target=/home/koyomi/lanraragi/database \
--mount type=bind,source=[YOUR_THUMBNAIL_DIRECTORY],target=/home/koyomi/lanraragi/thumb
difegue/lanraragi
```

{% hint style="info" %}
You can tell Docker to auto-restart the LRR container on boot by adding the `--restart always` flag to this command.

If you need to sideload plugins, consider adding an additional bind mount for the `/home/koyomi/lanraragi/lib/LANraragi/Plugin/Sideloaded` folder.
{% endhint %}

{% hint style="info" %}
Alternatively, if you prefer a `docker-compose.yml` file, use:

```yaml
---
services:
  lanraragi:
    image: difegue/lanraragi
    container_name: lanraragi
    ports:
      - "3000:3000"
    volumes:
      - [YOUR_CONTENT_DIRECTORY]:/home/koyomi/lanraragi/content
      - [YOUR_THUMBNAIL_DIRECTORY]:/home/koyomi/lanraragi/thumb
      - [YOUR_DATABASE_DIRECTORY]:/home/koyomi/lanraragi/database
    restart: unless-stopped
```

{% endhint %}

The content directory you have to specify in the command above will contain archives you either upload through the software or directly drop in, alongside generated thumbnails.\
The database directory houses the LANraragi database(As database.rdb), allowing you to hotswap containers without losing any data.

{% hint style="info" %}
You can also mount the database/thumbnail directories to dedicated Docker volumes:

```bash
docker volume create lrr-database
docker volume create lrr-thumbnails
docker run --name=lanraragi -p 3000:3000 \
--mount type=bind,source=[YOUR_CONTENT_DIRECTORY],target=/home/koyomi/lanraragi/content \
--mount source=lrr-database,target=/home/koyomi/lanraragi/database \
--mount source=lrr-thumbnails,target=/home/koyomi/lanraragi/thumb \
difegue/lanraragi
```

The volume can be reused when updating, so your database will still follow along even if the container is destroyed.\
You can always backup the database using LANraragi's internal tool.
{% endhint %}

Once your LANraragi container is loaded, you can access it at <http://localhost:3000> .\
You can use the following commands to stop/start/remove the container(Removing it won't delete the archive directory you specified) :

```bash
docker stop lanraragi
docker start lanraragi
docker rm lanraragi
```

[Tags](https://hub.docker.com/r/difegue/lanraragi/tags/) exist for major releases, so you can use those if you want to run another version:\
`docker run [yadda yadda] difegue/lanraragi:0.4.0`

{% hint style="danger" %}
If you're feeling **extra dangerous**, you can run the last files directly from the *dev* branch of the Git repo through the *nightly* tag:\
`docker run [zoinks] difegue/lanraragi:nightly`
{% endhint %}

## Platform-specific caveats

### Windows

If you're running on Windows, please check the syntax for mapping your content directory [here](https://docs.docker.com/docker-for-windows/#shared-drives).

Windows 7/8 users running the Legacy Docker toolbox will have to explicitly forward port 127.0.0.1:3000 from the host to the container in order to be able to access the app.

### Raspbian

If you're using **Raspbian**, it's likely you'll encounter installation issues like `s6-svscan: warning: unable to iopause: Operation not permitted` due to their outdated version of `libseccomp`.\
You can fix this by either adding `--security-opt seccomp=unconfined` to your Docker arguments(discouraged, allows LRR wider access to underlying OS), or by installing an up-to-date version of `libseccomp`:

```bash
wget http://ftp.debian.org/debian/pool/main/libs/libseccomp/libseccomp2_2.5.1-1~bpo10+1_armhf.deb
sudo dpkg -i libseccomp2_2.5.1-1~bpo10+1_armhf.deb
```

Regular versions of Debian shouldn't have this issue.

## Changing the port

Since Docker allows for port mapping, you can most of times map the default port of 3000 to another port on your host quickly.\
If you need something a bit more involved (like adding SSL), please check the Network Interfaces section for how to use the `LRR_NETWORK` environment variable.

{% content-ref url="../advanced-usage/network-interfaces" %}
[network-interfaces](https://sugoi.gitbook.io/lanraragi/advanced-usage/network-interfaces)
{% endcontent-ref %}

{% hint style="info" %}
The default healthchecks of the Docker container base themselves on port 3000.\
If you use the LRR\_NETWORK variable to change the outgoing port instead of Docker's port mapping, said healthchecks will fail.\
If you have to use the variable for SSL or the like, I recommend leaving the port in it to 3000 and doing your port mapping on the Docker side.
{% endhint %}

## Changing the user ID in case of permission issues

The container runs the software by default using the uid/gid provided by the LRR\_UID/LRR\_GID variables.\
If you don't specify said variables, the container will run under uid/gid 9001/9001.

This is good enough for most scenarios, but in case you need to run it as the current user, you can do the following: `docker run [wassup] -e LRR_UID=``id -u $USER`` -e LRR_GID=``id -g $USER`` difegue/lanraragi`

This uses `id` to automatically fetch your userid/groupid.

## Skip permission fixing

By default, LANraragi will change the permissions for all files and folders in the thumbnail and content folder during startup:

* Everything in the thumbnail folder will be owned by the provided UID/GID (9001/9001 by default)
* Files and folders in the thumbnail folder can be read and written by the owner
* Files and folders in the content folder are readable by the owner
* Folders in the thumbnail/content folder can be executed by the owner

While this can resolve some permission issues, but it will result in a longer start-up time for large library and may even prevent the container from starting on Docker Swarm.

To disable this behaviour, you can set the environment variable `LRR_AUTOFIX_PERMISSIONS=-1`, but make sure you have manually set the permissions beforehand.

## Updating

As Docker containers are immutable, you need to destroy your existing container and build a new one.

```bash
docker pull difegue/lanraragi
docker stop lanraragi
docker rm lanraragi
docker run --name=lanraragi -p 3000:3000 \
           --mount type=bind,source=[YOUR_CONTENT_DIRECTORY],target=/home/koyomi/lanraragi/content \
           --mount type=bind,source=[YOUR_DATABASE_DIRECTORY],target=/home/koyomi/lanraragi/database \
           --mount type=bind,source=[YOUR_THUMBNAIL_DIRECTORY],target=/home/koyomi/lanraragi/thumb
           difegue/lanraragi
```

As long as you use the same content/database directories(or volumes) as before, your data will still be there.

{% hint style="info" %}
If you update often, you might want to consider using docker-compose or [Portainer](https://portainer.io) to redeploy containers without entering the entire configuration every time.
{% endhint %}

## Building your own

The previous setup gets a working LANraragi container from the Docker Hub, but you can build your own bleeding edge version by executing `npm run docker-build` from a cloned Git repo.

This will use your cloned Git repo to build the image, modifications you might have made included.

Of course, this requires a Docker installation.\
If you're running WSL1, which can't run Docker natively, you can directly use the Docker for Windows executable with a simple symlink:

```bash
sudo ln -s '/mnt/c/Program Files/Docker/Docker/resources/bin/docker.exe' \
/usr/local/bin/docker
```