- 1 Introduction
- 2 Requirements
- 3 Versions of this guide
- 4 Prepare the system
- 5 Get the source
- 6 Configuration
- 7 Bundle
- 8 Database setup
- 9 Precompile assets
- 10 Start diaspora*
- 11 Backup
- 12 Further reading
You don't need to have your own pod to use diaspora*.
Running your own diaspora* server allows more control over your data, but it requires technical skills and time. Instead, you can register on one of the installations open to everyone that allow you to use diaspora* in a few clicks without any requirements
This guide will outline the procedure to get you set up with a production-ready installation of diaspora*.
Things to know
- The install is a bit complex, but we're here to help.
It's extremely helpful to have some experience in Linux/Unix server administration or Rails app deployment already. But don't worry, if you run into problems and need help, just visit us in our IRC channels on Freenode.
- Running a common setup will get you the most help, if you need it.
Most people in the community will have some experience running diaspora* with Unicorn as the app server using Nginx as outward-facing web server. Of course, you're free to run any other app server (Thin, Passenger...) or web server (Apache), but you might find it harder to get help if you run into unexpected troubles.
- diaspora* is developed utilizing latest web standards
Therefore UX is best with recent browsers, so please update your Firefox, Opera, Chrome or Safari to the newest version. We do not currently support any version of Internet Explorer, though we won't reject any contributions attempting to change that circumstance.
- diaspora* strongly recommends HTTPS
as we encrypt communication amongst servers and to the client browsers. You can get a free TLS certificate from Let's Encrypt.
Unfortunately, self-signed certificates or certificates issued by CACert won't work.
- We need your feedback
to constantly improve and update this guide. Have a look at How we communicate
- Do not run any of the commands you find in this guide as root (except if requested).
Just use your normal user - or even better - create a separate user for diaspora* (rationale).
- Memory: 1.5 GB
- Swap: 1 GB
- CPU: decent multicore
- Storage: The amount of hard disk space required largely depends on how many images you expect your users to upload.
It is possible to run a pod on a Raspberry Pi >= 2. However, this will be very slow and is not recommended for multi-user pods.
Over the course of this manual, you will install the following software if not already installed.
- Build tools - for compiling source packages
- Ruby - the Ruby programming language
- RubyGems - package manager for Ruby code libraries (like CPAN for Perl or PEAR for PHP)
- Bundler - gem management tool for Ruby projects
- MySQL or MariaDB or PostgreSQL - backend storage engine
- OpenSSL - encryption library.
- libcurl - multiprotocol file transfer library WARNING: Due to sidekiq longjmp error, you need at least curl 7.32
- ImageMagick - image processing library
- Git - version control system
- Redis - persistent key-value store
Please note that running diaspora* together with other applications that use Redis on the same machine can be dangerous. Only do that if you absolutely know what you are doing, and if you know how to change the Redis database for all of the applications.
Versions of this guide
Make sure to use the correct version of the guide, see below.
» Help me decide!
In Production mode, your pod is configured to deal with high load for everyday usage. This is recommended for a pod you want to actually use.
In Development mode, your pod is configured for development. This is recommended ony for development contributors of diaspora* which use the pod only locally for testing purposes.
The Database you choose is up to you. In our experience PostgreSQL achieves a better performance but needs more resources.
The current guide is for a production setup with PostgreSQL as database.
You can change them below:
Running mode: Production |
Database: | PostgreSQL |
Prepare the system
As root run:
yum install tar make automake gcc gcc-c++ git net-tools libcurl-devel libxml2-devel libffi-devel libxslt-devel wget redis ImageMagick nodejs postgresql-devel
Install the database
If you already have one skip this step.
As root run:
yum install postgresql-server postgresql-setup initdb systemctl enable postgresql systemctl start postgresql systemctl enable redis systemctl start redis
Creating a user for Diaspora
As root run:
adduser --disabled-login diaspora # If you already created a user and want to disable login for it: usermod -L diaspora # Add the new user to the wheel group so that you are able to access sudo functionality usermod -aG wheel diaspora # Then switch to it: su diaspora cd ~
The rest of the guide should happen under this user!
Creating a user for DB
If you like a separate DB user for your diaspora* installation, log in to your PostgreSQL server as the main user 'postgres'. One way to do this is:
sudo -u postgres psql
You need a user with the privilege to create databases.
CREATE USER diaspora WITH CREATEDB PASSWORD '<password>';
We recommend using Ruby Version Manager it will ensure you're always on the currently recommended Ruby version and cleanly separate your diaspora* installation from all other Ruby applications on your machine. If you opt for not using it ensure your Ruby version is at least 2.3.0, prior versions are incompatible. We currently recommend using the latest release of the 2.4 series.
As the user you want to run diaspora* under, that is not as root, run:
curl -L https://s.diaspora.software/1t | bash
and follow the instructions. If you get GPG signature problems, follow the instructions printed by the command. Running the 'gpg --recv-keys' command with 'sudo' should not be necessary. If those commands give you permission denied errors, change them to 640 for all files and 750 for all folders in the .gnupg folder.
Set up RVM
Ensure the following line is in your ~/.bashrc:
[[ -s "$HOME/.rvm/scripts/rvm" ]] && source "$HOME/.rvm/scripts/rvm"
Now run source ~/.bashrc in the terminal(s) you are using for this guide.
If you don't have sudo installed or your current user doesn't have the privileges to execute it, run:
rvm autolibs read-fail
The next command will check if all dependencies to build Ruby are installed. If these are not met, you will see a list of packages preceded by "Missing required packages:". As root install all the packages listed there for your OS. Then rerun the install command.
Ensure the currently recommend version of Ruby is installed:
rvm install 2.4
Get the source
It's time to download diaspora*! As your diaspora user run:
cd ~ git clone -b master https://github.com/diaspora/diaspora.git cd diaspora
Don't miss the cd diaspora, all coming commands expect to be run from that directory!
cp config/database.yml.example config/database.yml cp config/diaspora.yml.example config/diaspora.yml
Now open config/database.yml and config/diaspora.yml in your favorite text editor and carefully review them, they are extensively commented.
Important values in config/database.yml
- username Make sure to use the right username you created for diaspora.
- password Make sure to use the right password for the user your created for diaspora.
Important values in config/diaspora.yml
- environment.url: Set the public facing URL to your pod here, for example for https://pod.geraspora.de this would be https://pod.geraspora.de
- environment.certificate_authorities: You have to set this, one of the examples should fit. If the file in the example doesn't exist you're missing a package, in most cases it's named ca-certificates.
- server.rails_environment: You must set this to production. The server section is read by ./script/server and most alternative startup methods to setup the correct environment.
- environment.require_ssl: If for some reason you can't run your pod on HTTPS (we highly encourage you to do it!), set this to false to prevent a redirect from http:// to https://
You most likely have already a webserver running on port 80 (http) and 443 (https). It also should serve Diasporas static content and forward all other requests to Diaspora. Here are some example configurations to achieve that:
- Apache: https://gist.github.com/719014
- Nginx: https://gist.github.com/1355430
- Caddy: https://gist.github.com/oliof/57345a6596bb3a564f68d1482fb383f9
It's time to install the Ruby libraries required by diaspora*:
gem install bundler script/configure_bundler bin/bundle install --full-index
This takes quite a while. When it's finished, you should see a message similar to: Bundle complete! 137 Gemfile dependencies, 259 gems now installed. If that's not the case, you should seek for help on the mailing list or the IRC channel.
Running the manual gem install command shown in the error message can sometimes show a clearer error message if the bundle command fails.
Double check your config/database.yml looks right and run:
RAILS_ENV=production bundle exec rake db:create db:migrate
RAILS_ENV=production bin/rake assets:precompile
It's time to start diaspora*:
Your diaspora server is now running, either on a unix socket (current default) or on http port 3000. The listening method can be configured in diaspora.yml, search for '3000' or 'listen' to find the correct line.
You will likely need to install a reverse proxy (example on github for apache2) in order to get it to be served publicly. If you are new to running rails applications you may find the diaspora components page helpful for orientation.
You have to do backups of your pod data. If you lose your data, you won't be able to use the combination of your old username and old domain ever again. Make sure to store the backups on a different server, or at least on a different hard drive.
For details on how to do backups, see Pod data backup.
- Diasporas components explained
- Is there anybody out there? Establish connections with other pods
- Make yourself an admin
- Updating Diaspora
- Integrating other social networks
- Alternative startup methods (Passenger, Init script, Daemontools, ...)
- Asset hosting on S3