Home > Article > Backend Development > nginx+unicorn deploy ruby web environment
I recently took over a ruby web project. Since I have never been exposed to ruby and related web development before, and the documentation in the project is missing, I took some detours when deploying the project. Finally, I successfully deployed the project by referring to multiple ruby on rails development environment deployments and learning the relationship between gems/bundlers in ruby. Write it down here, it is a small summary, and it can also provide a reference for the deployment environment when you are just getting started with ruby web development.
Our project is based on padrino. If it is a rails-based project, the deployment should be similar.
1. Install ruby
Since there may be multiple ruby-related projects in the system, and different projects may need to be developed or tested for different ruby versions, multiple ruby versions may need to be installed in the system . This requires corresponding tools to manage existing versions and install new versions. I choose rbenv here, but you can also use the old rvm.git clone git://github.com/sstephenson/rbenv.git ~/.rbenv echo 'export PATH="$HOME/.rbenv/bin:$PATH"' >> ~/.bashrc echo 'eval "$(rbenv init -)"' >> ~/.bashrc exec $SHELLAdd ruby-build component
git clone git://github.com/sstephenson/ruby-build.git ~/.rbenv/plugins/ruby-build echo 'export PATH="$HOME/.rbenv/plugins/ruby-build/bin:$PATH"' >> ~/.bashrc exec $SHELLAdd gem-rehash component
git clone https://github.com/sstephenson/rbenv-gem-rehash.git ~/.rbenv/plugins/rbenv-gem-rehashInstall ruby (take ruby2.1.5 as an example)
wget -O ~/.rbenv/versions/ruby-2.1.5.tar.gz http://ruby.taobao.org/mirrors/ruby/2.1/ruby-2.1.5.tar.gz env RUBY_BUILD_MIRROR_URL=file:///home/ziven/.rbenv/versions/ruby-2.1.5.tar.gz# ~/.rbenv/bin/rbenv install 2.1.5where wget is to download the corresponding ruby file, here is ruby-2.1.5.tar.gz
RUBY_BUILD_MIRROR_URLis used to specify the installed ruby image path, here the local path is used directly. Note that the local path should be ended with #. The version following the rbenv install command should correspond to the ruby version to be installed. For example, the installed version may have a corresponding patch number (for example: 1.9.3-p223). This should also be written.
Because we generally use bundler to manage the gems needed for the project, bundler needs to be installed here.
Considering the speed of connecting ruby mirrors in China, we changed the gem source to taobao source.gem sources --remove https://rubygems.org/ gem sources -a https://ruby.taobao.org/After executing gem sources -l, there should be only one source: https://ruby.taobao.org/
gem install bundler
Here we assume that after cloning Place the source code to /opt/projectA
Execute in the root directory of the project:
bundle installIt should be noted here that the commands executed may be slightly different depending on the situation. For example, when deploying in a production environment, you may need to execute
bundle install --deployment --without development test. When deploying like this, the relevant dependent gems are installed under projectA/vendor/bundle. It is especially suitable for use when different projects depend on different versions of the same gem.
Since this project is based on padrino, it needs to be executed:
padrino rake --envir db:migrateIt should be noted that depending on the situation, you may need to use bundle to execute this command:
bundle exec padrino rake --envir db:migrateAfter the database migration is successful, you need to create an amdin account. Padrino completes the creation of the admin account in db/seeds.rb by default. So execute here:
bundle exec padrino rake --envir seedwill prompt you to enter the email address and password of the admin account during the execution process
You may need the corresponding unicorn startup script here, which can be written by referring to ruby-china's unicorn.rb. Omitted here.
Start unicorn server:bundle exec unicorn -E production -c unicorn.rb -Dso padrino server has started normally. If unicorn.rb is listening on the tcp port, then you should be able to access the corresponding URL at this time. If you use the method of monitoring unix socket, you should also use nginx (or apache) as a reverse proxy.
nginx can be installed in the same way as general Linux software. It is recommended to use the source code installation method. For the sake of simplicity here, we will temporarily install it using the package manager installation method:
sudo apt-get install nginx-full
upstream myserver { server unix:/tmp/projectA.sock fail_timeout=0; } server { listen 7788; server_name localhost; client_max_body_size 10M; error_page 500 502 503 504 /50x.html; location = /50x.html { root /usr/share/nginx/html; } root /opt/projectA/web/public; try_files $uri $uri.html @app; location @app { proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; proxy_set_header Host $http_host; proxy_redirect off; proxy_pass http://myserver; } }Here, the request for port 7788 is forwarded to
unix:/tmp/projectA.sockfor processing. That is, it is forwarded to our web server just now for processing. At this point, the server is running normally. Other auxiliary functions can be added as needed. For example, nginx may require an SSL certificate for authentication.