capistrano-postgresql 3.0.1 → 4.0.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 8268ca43327dee94bcb0286b28830e5835fdf867
-  data.tar.gz: 277c459ccd2660f3cc4daaecd638f6d79d43e10e
+  metadata.gz: a4979d4dae2bd808a75c9656d8206de81940ed8a
+  data.tar.gz: 985c1c832fcde7f17e129e196e1a4d41ee19420f
 SHA512:
-  metadata.gz: 8ab557923a1b61c8979920759870153812e2e7aada895679fdddcc0e2ab2cc2b8b4713b709912c99b37faecb0fa12a8bb35985cfe6ad13ab4a6aee7a2341af93
-  data.tar.gz: 1cc9cc010a2ab99fd7f5efc61256ac27f6e123e001b71c26a1c49ac1fb253f5897421e3dc9f684e6ef4151636c8f3e95900722e7b53be7ac82c707153e00d6cf
+  metadata.gz: dc3160801f01359dfe35aad44085910b8056ddf09be8e194caff6b0dbb70e0427f469a6223e288c32a10d97085d5d67f55f0cd13237428ab753df08b0bb4da79
+  data.tar.gz: c4cf87e7bff7d00e1163f50cdd54e9a9bca1ad55b0be1c9bd2bb96edf9a289006bc0a3dc7a4e0058682e57cc01e123caa8469070abff0380ed80c8c68b6a2bbb

data/CHANGELOG.md

@@ -1,5 +1,19 @@
+# Changelog
+
+### master
+
+### v4.0.0, 2014-10-06
+- enable setting DB environment with `rails_env` option. If `rails_env` is not
+  set, `stage` option is used as until now. (@bruno-)
+- create a task that helps with the upgrade to gem version 4 (@bruno-)
+- optionally create a hstore extension on the server (@rhomeister)
+- `database.yml` is now copied to all release_roles (@rhomeister)
+- introduce archetype `database.yml` that is stored on primary `db` node and
+  is copied to all `release` nodes on `setup` task (@rhomeister)
+
 ### v3.0.0, 2014-04-11
 - all the work is moved to the `setup` task
+
 ### v2.0.0, 2014-03-30
 - shorten variable names: postgresql -> pg
 - better helpers module separation
@@ -7,5 +21,6 @@
 - lots of code improvements
 - less layered and simpler code
 - readme updates
+
 ### v1.0.0, 2014-03-19
 - all the v1.0.0 features

data/README.md

@@ -15,17 +15,22 @@ Here are the specific things this plugin does for your capistrano deployment
 process:
 
 * creates a new PostgreSQL database and database user on the server
-* generates and populates `database.yml` file
+* generates and populates `database.yml` file on all release nodes
   (no need to ssh to the server and do this manually!)
 * zero-config
+* support for multi-server setup: separate `db` and `app` nodes (from version 4.0)
+
+**Note**: gem version 4 introduces some breaking changes. If you installed gem
+version 3 or below you might want to follow the
+[upgrade instructions](https://github.com/capistrano-plugins/capistrano-postgresql/wiki/Upgrade-instructions-for-gem-version-4.0).
 
 ### Installation
 
 Put the following in your application's `Gemfile`:
 
     group :development do
-      gem 'capistrano', '~> 3.1'
-      gem 'capistrano-postgresql', '~> 3.0'
+      gem 'capistrano', '~> 3.2.0'
+      gem 'capistrano-postgresql', '~> 4.0.0'
     end
 
 Then:
@@ -42,139 +47,43 @@ the following in `Capfile` file:
 Make sure the `deploy_to` path exists and has the right privileges on the
 server (i.e. `/var/www/myapp`).<br/>
 Or just install
-[capistrano-safe-deploy-to](https://github.com/bruno-/capistrano-safe-deploy-to)
+[capistrano-safe-deploy-to](https://github.com/capistrano-plugins/capistrano-safe-deploy-to)
 plugin and don't think about it.
 
-To setup the server, run:
+To setup the server(s), run:
 
     $ bundle exec cap production setup
 
-Easy, right?
-
-Check below to see what happens in the background.
-
-### How it works
-
-Check here for the full capistrano deployment flow
-[http://capistranorb.com/documentation/getting-started/flow/](http://capistranorb.com/documentation/getting-started/flow/).
-
-The following tasks run during the `setup` task:
-
-* `postgresql:create_db_user`<br/>
-creates a postgresql user. Password for the user is automatically generated and
-used in the next steps.
-* `postgresql:create_database`<br/>
-creates database for your app.
-* `postgresql:generate_database_yml`<br/>
-creates a `database.yml` file and copies it to
-`#{shared_path}/config/database.yml` on the server.
-
-Also, the plugin ensures `config/database.yml` is symlinked from `shared_path`.
-The above tasks are all you need for getting a Rails app to work with PostgreSQL.
-
 ### Gotchas
 
 Be sure to remove `config/database.yml` from your application's version control.
 
-### Configuration
-
-This plugin should just work with no configuration whatsoever. However,
-configuration is possible. Put all your configs in capistrano stage files i.e.
-`config/deploy/production.rb`.
-
-Here's the list of options and the defaults for each option:
-
-* `set :pg_database`<br/>
-Name of the database for your app. Defaults to `#{application}_#{stage}`,
-example: `myface_production`.
-
-* `set :pg_user`<br/>
-Name of the database user. Defaults to whatever is set for `pg_database`
-option.
-
-* `set :pg_password`<br/>
-Password for the database user. By default this option is not set and a
-**new random password** is generated each time you create a new database.<br/>
-If you set this option to `"some_secure_password"` - that will be the db user's
-password. Keep in mind that having a hardcoded password in `deploy.rb` (or
-anywhere in version control) is a bad practice.<br/>
-I recommend sticking to the default and generating a new secure and random
-password each time a db user is generated. That way you don't have to worry
-about it or try to remember it.
-
-* `set :pg_ask_for_password`<br/>
-Default `false`. Set this option to `true` if you want to be prompted for the
-password when database user is created. This is safer than setting the password
-via `pg_password`. The downside is you have to choose and remember
-yet another fricking password.<br/>
-`pg_password` option has precedence. If it is set,
-`pg_ask_for_password` is ignored.
-
-* `set :pg_system_user`<br/>
-Default `postgres`. Set this option to the user that owns the postgres process
-on your system. Normally the default is fine, but for instance on FreeBSD the
-default prostgres user is `pgsql`.
+### How it works
 
-* `set :pg_system_db`<br/>
-Default `postgres`. Set this if the system database don't have the standard name.
-Usually there should be no reason to change this from the default.
+[How the plugin works](https://github.com/capistrano-plugins/capistrano-postgresql/wiki/How-it-works)
+wiki page contains a list of actions the plugin executes.
 
-`database.yml` template-only settings:
+Read it only if you want to learn more about the plugin internals.
 
-* `set :pg_pool`<br/>
-Pool config in `database.yml` template. Defaults to `5`.
+### Configuration
 
-* `set :pg_host`<br/>
-`hostname` config in `database.yml` template. Defaults to `localhost`.
+A full
+[list of configuration options](https://github.com/capistrano-plugins/capistrano-postgresql/wiki/Configuration-options).
 
-* `set :pg_encoding`<br/>
-`encoding` config in `database.yml` template. Defaults to `unicode`.
+The list can be overwhelming so consult it only if you're looking for something
+specific.
 
 ### Customizing the `database.yml` template
 
-This is the default `database.yml` template that gets copied to the capistrano
-shared directory on the server:
-
-```yml
-<%= fetch :stage %>:
-  adapter: postgresql
-  encoding: <%= pg_encoding %>
-  database: <%= pg_database %>
-  pool: <%= pg_pool %>
-  username: <%= pg_user %>
-  password: '<%= pg_password %>'
-  host: <%= pg_host %>
-```
-
-If for any reason you want to edit or tweak this template, you can copy it to
-`config/deploy/templates/postgresql.yml.erb` with this command:
-
-    $ bundle exec rails g capistrano:postgresql:template
-
-After you edit this newly created file in your repo, it will be used as a
-template for `database.yml` on the server.
-
-You can configure the template location. For example:
-`set :pg_templates_path, "config"` and the template will be copied to
-`config/postgresql.yml.erb`.
+[Wiki page about the database.yml format](https://github.com/capistrano-plugins/capistrano-postgresql/wiki/Customizing-the-database.yml-template).
 
 ### More Capistrano automation?
 
-If you'd like to streamline your Capistrano deploys, you might want to check
-these zero-configuration, plug-n-play plugins:
-
-- [capistrano-unicorn-nginx](https://github.com/bruno-/capistrano-unicorn-nginx)<br/>
-no-configuration unicorn and nginx setup with sensible defaults
-- [capistrano-rbenv-install](https://github.com/bruno-/capistrano-rbenv-install)<br/>
-would you like Capistrano to install rubies for you?
-- [capistrano-safe-deploy-to](https://github.com/bruno-/capistrano-safe-deploy-to)<br/>
-if you're annoyed that Capistrano does **not** create a deployment path for the
-app on the server (default `/var/www/myapp`), this is what you need!
+Check out [capistrano-plugins](https://github.com/capistrano-plugins) github org.
 
 ### Contributing and bug reports
 
-Contributions and improvements are very welcome. Just open a pull request and
-I'll look it up shortly.
+Contributions and improvements are very welcome.
 
 If something is not working for you, or you find a bug please report it.
 

data/lib/capistrano/postgresql/helper_methods.rb

@@ -15,10 +15,16 @@ module Capistrano
         StringIO.new ERB.new(File.read(config_file)).result(binding)
       end
 
+      # location of database.yml file on clients
       def database_yml_file
         shared_path.join('config/database.yml')
       end
 
+      # location of archetypical database.yml file created on primary db role when user and
+      # database are first created
+      def archetype_database_yml_file
+        deploy_path.join("db/database.yml")
+      end
     end
   end
 end

data/lib/capistrano/postgresql/psql_helpers.rb

@@ -4,7 +4,12 @@ module Capistrano
 
       # returns true or false depending on the remote command exit status
       def psql(*args)
-        test :sudo, "-u #{fetch(:pg_system_user)} psql -d #{fetch(:pg_system_db)}", *args
+        psql_on_db(fetch(:pg_system_db), *args)
+      end
+
+      # Runs psql on the application database
+      def psql_on_app_db(*args)
+        psql_on_db(fetch(:pg_database), *args)
       end
 
       def db_user_exists?(name)
@@ -15,6 +20,10 @@ module Capistrano
         psql '-tAc', %Q{"SELECT 1 FROM pg_database WHERE datname='#{db_name}';" | grep -q 1}
       end
 
+      private
+      def psql_on_db(db_name, *args)
+        test :sudo, "-u #{fetch(:pg_system_user)} psql -d #{db_name}", *args
+      end
     end
   end
 end

data/lib/capistrano/postgresql/version.rb

@@ -1,5 +1,5 @@
 module Capistrano
   module Postgresql
-    VERSION = "3.0.1"
+    VERSION = "4.0.0"
   end
 end

data/lib/capistrano/tasks/postgresql.rake

@@ -14,8 +14,10 @@ namespace :load do
     set :pg_password, -> { ask_for_or_generate_password }
     set :pg_system_user, 'postgres'
     set :pg_system_db, 'postgres'
+    set :pg_use_hstore, false
     # template only settings
     set :pg_templates_path, 'config/deploy/templates'
+    set :pg_env, -> { fetch(:rails_env) || fetch(:stage) }
     set :pg_pool, 5
     set :pg_encoding, 'unicode'
     set :pg_host, 'localhost'
@@ -24,19 +26,42 @@ end
 
 namespace :postgresql do
 
+  desc 'Steps to upgrade the gem to version 4.0'
+  task :upgrade4 do
+    on roles :db do
+      execute :mkdir, '-pv', File.dirname(archetype_database_yml_file)
+      execute :cp, database_yml_file, archetype_database_yml_file
+    end
+  end
+
   # undocumented, for a reason: drops database. Use with care!
   task :remove_all do
-    on roles :app do
+    on release_roles :all do
       if test "[ -e #{database_yml_file} ]"
         execute :rm, database_yml_file
       end
     end
+
+    on primary :db do
+      if test "[ -e #{archetype_database_yml_file} ]"
+        execute :rm, archetype_database_yml_file
+      end
+    end
+
     on roles :db do
       psql '-c', %Q{"DROP database #{fetch(:pg_database)};"}
       psql '-c', %Q{"DROP user #{fetch(:pg_user)};"}
     end
   end
 
+  desc "Add the hstore extension to postgresql"
+  task :add_hstore do
+    next unless fetch(:pg_use_hstore)
+    on roles :db do
+      psql_on_app_db '-c', %Q{"CREATE EXTENSION IF NOT EXISTS hstore;"}
+    end
+  end
+
   desc 'Create DB user'
   task :create_db_user do
     on roles :db do
@@ -59,12 +84,29 @@ namespace :postgresql do
     end
   end
 
-  desc 'Generate database.yml'
+  # This task creates the archetype database.yml file on the primary db server. This is done once when a
+  # new DB user is created.
+  desc 'Generate database.yml archetype'
+  task :generate_database_yml_archetype do
+    on roles :db, primary: true do
+      next if test "[ -e #{archetype_database_yml_file} ]"
+      execute :mkdir, '-pv', File.dirname(archetype_database_yml_file)
+      upload! pg_template('postgresql.yml.erb'), archetype_database_yml_file
+    end
+  end
+
+  # This task copies the archetype database file on the primary db server to all clients. This is done on
+  # every setup, to ensure new servers get a copy as well.
+  desc 'Copy archetype database.yml from primary db server to clients'
   task :generate_database_yml do
-    on roles :app do
-      next if test "[ -e #{database_yml_file} ]"
+    database_yml_contents = nil
+    on primary :db do
+      database_yml_contents = download! archetype_database_yml_file
+    end
+
+    on release_roles :all do
       execute :mkdir, '-pv', shared_path.join('config')
-      upload! pg_template('postgresql.yml.erb'), database_yml_file
+      upload! StringIO.new(database_yml_contents), database_yml_file
     end
   end
 
@@ -73,11 +115,18 @@ namespace :postgresql do
   end
 
   after 'deploy:started', 'postgresql:database_yml_symlink'
+
+  desc 'Postgresql setup tasks'
+  task :setup do
+    invoke "postgresql:create_db_user"
+    invoke "postgresql:create_database"
+    invoke 'postgresql:add_hstore'
+    invoke "postgresql:generate_database_yml_archetype"
+    invoke "postgresql:generate_database_yml"
+  end
 end
 
 desc 'Server setup tasks'
 task :setup do
-  invoke 'postgresql:create_db_user'
-  invoke 'postgresql:create_database'
-  invoke 'postgresql:generate_database_yml'
+  invoke "postgresql:setup"
 end

data/lib/generators/capistrano/postgresql/templates/postgresql.yml.erb

@@ -1,4 +1,4 @@
-<%= fetch :stage %>:
+<%= fetch :pg_env %>:
   adapter: postgresql
   encoding: <%= fetch :pg_encoding %>
   database: <%= fetch :pg_database %>

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: capistrano-postgresql
 version: !ruby/object:Gem::Version
-  version: 3.0.1
+  version: 4.0.0
 platform: ruby
 authors:
 - Bruno Sutic
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-07-20 00:00:00.000000000 Z
+date: 2014-10-06 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: capistrano