A docker-compose based deployment intended to bring Seafile on any ARMv7/ARM64 board with little effort. MariaDB/SQLite and a reverse-proxy with automated TLS certificates renewal are included.
Don't want automated TLS certs? See the No SWAG section below.
This environment has been set up using the following images and packages (which you should glance at for further documentation):
- Base Docker image
- linuxserver/mariadb as database server
- linuxserver/swag as reverse-proxy with cerbot support for Let's Encrypt.
Some other useful docs:
No guarantees of any kind are provided by using this environment.
A working docker-compose environment, a domain name associated with your server and the TCP ports 80 and 443 correctly forwarded.
Clone this repository somewhere on your server:
$ git clone https://github.com/ChatDeBlofeld/seafile-arm-docker
$ cd seafile-arm-docker
The topology provided should easily fulfill basic use cases. For a finer configuration, see the docs mentioned above.
Note: the
compose.shscript is a shortcut to thedocker-composecmd, it assembles the right files for your configuration (mainly). You should it instead ofdocker-composejust to have things working.
Copy the .env.example file:
$ cp .env.example .env
Then edit the dotenv with your favorite editor and take care at least of the following topics:
- Set
DBMSto 0 (SQLite), 1 (MariaDB) or 2 (MySQL) to select your DBMS. - Set your domain at the
HOSTvariable. - All variables mentioning credentials (email, passwords) must be updated for obvious security reasons.
- All volumes are mapped to the current directory by default, feel free to remap them to the most appropriate place for you, for example an external drive.
If compose V2 is installed, you can take advantage of the new config command to generate a compose file which can be used with the more traditional docker compose command instead of the launch script.
$ ./compose.sh config > docker-compose.ymlFirst generate the certificates by running the service once:
$ ./compose.sh up reverse-proxy
If everything went right, stop the container and copy the file nginx/seafile.conf to /path/to/swag/volume/nginx/site-confs and set your domain at the mentioned sections.
You may want to see the following resources for more information about this configuration:
- https://manual.seafile.com/deploy/deploy_with_nginx/
- https://manual.seafile.com/deploy/https_with_nginx
Alternatively, you may want not to use the SWAG container for testing and whatever reasons. Then:
- Set
NOSWAG=1in your dotenv. - Edit the file
seafile.noswag.confand set your local ip at the mentioned section.
Then depending on your use case configuration may differ.
In addition to the former, set USE_HTTPS=false in your dotenv. The application will be listening on port 80, if you want to initialize with another port, both NOSWAG_PORT and PORT variables have to be edited and match.
Simply run:
$ ./compose.sh up -d
You should now be able to access http(s)://your.domain and log in with your admin account.
Note: after the first run, all the credentials related environment variables can be removed from the dotenv file.
Currently there's no breaking changes between the images, so the update is straightforward:
$ ./compose.sh pull seafile
$ ./compose.sh up -d
Note: rollback will likely fail for major updates
Like the seafile service, a pull of the new image should do the job.
The Nginx configuration may be updated sometimes, then it has to be downloaded from this repository and copied to the right folder again.
For all default files provided by the swag container, see this updating procedure.
Set daemon to False in gunicorn.conf.py and restart the Seafile service to grab more information about a failed start.