Cheat Sheet CAPISTRANO 1.2.

0
Helper Methods config/deploy.rb Directory structure Shell commands
run Setting and using variables releases
Takes a single string identifying any valid shell 1 set :application, "flipper" Installation
command to execute 2 set :user, "homersimpson" 20050725121411 gem install capistrano
run <<-CMD 3 puts "The application name is 20050801090107
if [[ -d #{release_path}/status.txt #{application}" 20050802231414 Add your application to Capistrano (capistranize)
]]; then 4 puts "The user is #{user}" cap --apply-to /path/to/your/app
cat #{release_path}/status.txt ...
Redefining the restart task 20050824141402 YourApplicationName
fi
CMD 1 desc "This task restarts the web
server" Rakefile Execute the setup task
sudo 2 task :restart, :roles => app rake remote:exec ACTION=setup
Exactly like the run command, except that it :app do config
executes the command via sudo 3 sudo "apachectl graceful" Execute the cold_deploy task
sudo "apachectl graceful" 4 end db
lib rake remote:exec ACTION=cold_deploy
put Defining tasks
Lets you transfer data to a file on the remote 1 task :hello_world do log-->/shared/log Deploy your application
host. Takes two parameters: a string containing 2 run "echo Hello, $HOSTNAME" public rake deploy
the data to transfer, and the name of the file to 3 end system-->/shared/system
receive the data on each remote host Rollback a release from production
put(File.read('templates/database.yml' 1 task :hello_world, :roles => script
[:db, :app] do test rake rollback
),"#{release_path}/config/database.yml
",:mode => 0444) 2 puts "calling hello_world..." vendor Execute the disable_web task
delete 3 hello_world
4 end shared rake remote_exec ACTION=disable_web \
A convenience for executing rm via run. It just log UNTIL="tomorrow morning" \
attempts to do an rm -f on the remote Transactions
server(s), for the named file. To do a recursive 1 task :cold_deploy do system REASON="vital upgrade"
delete, pass :recursive => true 2 transaction do current-->/releases/20050824141402
delete "#{release_path}/certs", 3 task_one_here Using the invoke task
:recursive => true 4 task_two_here rake remote_exec ACTION=invoke \
on_rollback 5 end COMMAND="svn up
6 task_three_not_in_transaction
Allows a task to specify a callback to use if that
task raises an exception when invoked inside of 7 end Capistrano Shell /u/apps/flipper/current/app/views" \
ROLES=app
a transaction Capturing output with run
task :update_code do 1 run "sudo ls -la" do |channel, Start the interactive Capistrano shell
on_rollback { delete release_path, stream, data| cap -v shell
:recursive => true } 2 if data =~ /^Password:/
... 3 logger.info "#{channel[:host]} Execute Capistrano tasks Roles
end asked for password" !deploy
render 4 channel.send_data "mypass\n" 1 role :web , "www.capistrano.com"
5 end !update_code symlink 2 role :app , "app1.capistrano.com",
An interface for rendering ERb* templates and 6 end
returning the result. If you pass a string to !setup deploy "app2.capistrano.com"
render, it is interpreted as the name of a Parsing & saving an ERb template with render on app2.foo.com !setup 3 role :db , "master.capistrano.com",
template file with .rhtml appended, relative 1 buffer = render(:template => with app,db !setup deploy :primary => true
either to the current directory, or to <<EXAMPLE_TEMPLATE)
capistrano/recipes/templates to be 2 This template will be rendered 4 role :db , "slave.capistrano.com"
rendered. If you don’t want to render a file, but replacing variables 5 role :spare , "genghis.capistrano.com
instead have a string containing an ERb <%= like_this_variable =>
template that you want to render, use the with their values.  standard, predefined roles
second example below *Embedded Ruby 3 EXAMPLE_TEMPLATE  user-defined roles
render "maintenance" 4 put buffer,
render :template => "Hello <%= target "path/to/save/file.txt",
5 :mode => 0755
%>", :target => "world"
Standard tasks
Pre-defined variables cleanup cleans up the releases directory, leaving the five most recent releases
:application (required) The name of your application. Used to build other values, like the cold_deploy used when deploying an application for the first time. Starts the
deployment directory. application’s spinner (via the spinner task) and then does a normal
deploy
:repository (required) The location of your code’s scm repository
deploy updates all the code on your server (via update_code and symlink
:gateway nil The address of the server to use as a gateway. If given, all other tasks), then restarts the FastCGI listeners on the application servers (via
connections will be tunneled through this server. the restart task).
:user (current user) The name of the user to use when logging into the remote host(s) diff_from_last_ prints the difference between what was last deployed, and what is
:password (prompted) The password to use for logging into the remote host(s). deploy currently in your repository
:deploy_to “/u/apps/#{app The root of the directory tree on the remote host(s) that the disable_web puts up a static maintenance page that is displayed to visitors
lication}” application should be deployed to enable_web removes the maintenance page
:version_dir “releases” The directory under deploy_to that should contain each invoke allows you to send commands directly
deployed revision
migrate changes to the directory of your current release (as indicated by the
:current_dir “current” The name to use (relative to deploy_to) for the symlink that current symlink), and runs rake RAILS_ENV=production migrate
points at the current release
restart restarts all FastCGI listeners for your application by calling the reaper
:shared_dir “shared” The name of the directory under deploy_to that will contain command without arguments. Only executed on :app servers
directories and files to be shared between all releases
rollback rolls your application back to the previously deployed version
:revision (latest revision) This specifies the revision you want to check out on the remote
machines rollback_code determines the previous release , updates the current symlink to point
to that, and then deletes the latest release
:scm :subversion The source control module to use. Current supported are
:subversion, :cvs, :darcs setup Creates and chmods the directory tree:
releases_path directory 0775
:svn, :cvs, (path) The location on the remote host of the source control executable shared_path directory
:darcs shared_path/system 0775
:checkout “co” The subversion operation to use when checking out code on the shared_path/log 0777
remote host. Can be set to “export” show_tasks inspect the existing tasks and display them to standard out in
:ssh_options Hash.new Hash of additional options passed to the SSH connection routine. alphabetical order, along with their descriptions
This lets you set (among other things) a non-standard port to spinner starts the spinner process for your application
connect on (ssh_options[:port] = 2345)
symlink updates the current symlink to the latest deployed version of the code
:use_sudo true Whether or not tasks that can use sudo, ought to use sudo. In a
update_code Checks out your source code, deletes the log and public/system
shared environment, this is typically not desirable (or possible),
and in that case you should set this variable to false directories in your new release, symlinks log to
#{shared_path}/log, symlinks public/system to
:sudo sets the path to sudo #{shared_path}/system
task already has a defined on_rollback handler when using transactions

www.dizzy.co.uk/cheatsheets This work is licensed under the Creative Commons