Go to file
sn0w 8188426e58 Update README.md 2018-06-04 12:28:31 +00:00
custom.d Update structure 2018-04-08 22:53:33 +02:00
.dockerignore Update structure 2018-04-08 22:53:33 +02:00
.env.dist Revert PLEROMA_PORT 2018-06-04 01:51:25 +02:00
.gitignore Update structure 2018-04-08 22:53:33 +02:00
Dockerfile Introduce new Apache and NGINX mode. Enhance docs. 2018-06-04 01:27:03 +02:00
README.md Update README.md 2018-06-04 12:28:31 +00:00
docker-compose.m4 Revert PLEROMA_PORT 2018-06-04 01:51:25 +02:00
docker-config.exs Revert PLEROMA_PORT 2018-06-04 01:51:25 +02:00
entrypoint.ash Improve precompilation 2018-04-09 12:17:46 +02:00
initdb.sql Add pg_trgm extension 2018-06-04 01:58:02 +02:00
pleroma Update pleroma script 2018-06-04 01:31:52 +02:00


Pleroma-Docker (Unofficial)

Pleroma is a selfhosted social network that uses OStatus/ActivityPub.

This repository dockerizes it for easier deployment.

#include <public_domain.h>
#include <std_disclaimer.h>

 * This repository comes with ABSOLUTELY NO WARRANTY
 * I am not responsible for burning servers, angry users, fedi drama,
 * thermonuclear war, or you getting fired because your boss saw your
 * NSFW posts. Please do some research if you have any concerns about included
 * features or the software used by this script before using it.
 * You are choosing to use this setup, and if you point the finger at me for
 * messing up your instance, I will laugh at you.


  • 100% generic
  • Everything is customizable
  • Zero special host dependencies
  • Configuration is not compile-time
  • "It just works"


If this setup is a bit overwhelming there are a lot of other great dockerfiles or guides from the community. A few are linked below. This list is not exhaustive and not ordered.

Installing Pleroma

  • Clone this repository
  • Copy .env.dist to .env
  • Edit .env (see "Configuring Pleroma" section below)
  • Run ./pleroma build and ./pleroma start
  • Profit!

Updating Pleroma

Just run ./pleroma build again and ./pleroma start afterwards.

You don't need to shutdown pleroma while compiling the new release.

Every time you run ./pleroma build the script will fetch all upstream changes and checkout PLEROMA_VERSION. This means that setting PLEROMA_VERSION to a branch enables rolling-release updates while setting it to a tag or commit-hash pins the version.

Customizing Pleroma

Just add your customizations (and their folder structure) to custom.d.
They will be copied (not mounted) into the right place when the container starts.
You can even replace/patch pleroma's code with this, because the project is recompiled at startup.

In general: Prepending custom.d/ to pleroma's customization guides should work all the time.
Check them out in the official pleroma wiki.

For example: A custom thumbnail now goes into custom.d/priv/static/instance/thumbnail.jpeg instead of priv/static/instance/thumbnail.jpeg.

Configuring Pleroma

pleroma-docker tries to stay out of your way as much as possible while providing a good experience for both you and your users. It thus supports multiple "operation modes" and quite some config variables which you can mix and match.

This guide will explain some of the tricky .env file parts as detailed as possible (but you should still read the comments in there).

Since this setup injects code into pleroma that moves it's configuration into the environment (ref "The Twelve-Factor App"), the built image is 100% reusable and can be shared/replicated across multiple hosts. To do that just run ./pleroma build as usual and then tag your image to whatever you want. Just make sure to start the replicated container with env_file: or all required -e pairs.

Storing Data

Currently all data is stored in subfolders of DOCKER_DATADIR which will be bind-mounted into the container by docker.

We'll evaluate named volumes as an option in the future but they're currently not supported.


Values: true / false

By default pleroma-docker deploys a postgresql container and links it to pleroma's container as a zero-config data store. If you already have a postgres database or want to host postgres on a physically different machine set this value to false. Make sure to set the POSTGRES_* variables when doing that.

Reverse Proxy (SCRIPT_USE_PROXY)

Values: traefik / nginx / manual

Pleroma is usually run behind a reverse-proxy. Pleroma-docker gives you multiple options here.


In traefik-mode we will generate a pleroma container with traefik labels. These will be picked up at runtime to dynamically create a reverse-proxy configuration. This should 'just work' if watch=true and exposedByDefault=false are set in the [docker] section of your traefik.conf. SSL will also 'just work' once you add a matching [[acme.domains]] entry.


In nginx-mode we will generate a bare nginx container that is linked to the pleroma container. The nginx container is absolutely unmodified and expects to be configured by you. The nginx file in Pleroma's Repository is a good starting point.

We will mount your configs like this:

custom.d/server.nginx -> /etc/nginx/nginx.conf
custom.d/vhost.nginx -> /etc/nginx/conf.d/pleroma.conf

To reach your pleroma container from inside nginx use proxy_pass http://pleroma:4000;.

Set SCRIPT_PORT_HTTP and SCRIPT_PORT_HTTPS to the ports you want to listen on. Specify the ip to bind to in SCRIPT_BIND_IP. These values are required.

The container only listens on SCRIPT_PORT_HTTPS if SCRIPT_ENABLE_SSL is true.

Apache / httpd

Just like nginx-mode this starts an unmodified apache server that expects to be configured by you. Again Pleroma's Config is a good starting point.

We will mount your configs like this:

custom.d/server.httpd -> /usr/local/apache2/conf/httpd.conf
custom.d/vhost.httpd -> /usr/local/apache2/conf/extra/httpd-vhosts.conf

To reach your pleroma container from inside apache use ProxyPass [loc] http://pleroma:4000/.


The container only listens on SCRIPT_PORT_HTTPS if SCRIPT_ENABLE_SSL is true.


In manual mode we do not create any reverse proxy for you. You'll have to figure something out on your own.

This mode also doesn't bind to any IP or port. You'll have to forward something to the container's IP.


Values: true / false

If you want to use SSL with your Apache or NGINX containers you'll need a certificate. Certificates need to be placed into custom.d and will be bind-mounted into the server's container at runtime.

We will mount your certs like this:

custom.d/ssl.crt -> /ssl/ssl.crt
custom.d/ssl.key -> /ssl/ssl.key

You can reference them in Apache like this:

<VirtualHost *:443>
    SSLEngine on
    SSLCertificateFile "/ssl/ssl.crt"
    SSLCertificateKeyFile "/ssl/ssl.key"

And in NGINX like this:

listen 443 ssl;
ssl_certificate     /ssl/ssl.crt;
ssl_certificate_key /ssl/ssl.key;

In traefik-mode and manual-mode these files and the SCRIPT_LOCAL_SSL value are ignored.