Bump version for 1.4.0

Pass around Roles instead of Strings

.github/workflows/ci.yml

@@ -12,17 +12,29 @@ jobs:
           - "2.7"
           - "3.1"
           - "3.2"
+          - "3.3"
         gemfile:
           - Gemfile
+          - gemfiles/ruby_2.7.gemfile
           - gemfiles/rails_edge.gemfile
-        continue-on-error: [false]
+        exclude:
+          - ruby-version: "2.7"
+            gemfile: Gemfile
+          - ruby-version: "2.7"
+            gemfile: gemfiles/rails_edge.gemfile
+          - ruby-version: "3.1"
+            gemfile: gemfiles/ruby_2.7.gemfile
+          - ruby-version: "3.2"
+            gemfile: gemfiles/ruby_2.7.gemfile
+          - ruby-version: "3.3"
+            gemfile: gemfiles/ruby_2.7.gemfile
     name: ${{ format('Tests (Ruby {0})', matrix.ruby-version) }}
     runs-on: ubuntu-latest
-    continue-on-error: ${{ matrix.continue-on-error }}
+    continue-on-error: true
     env:
       BUNDLE_GEMFILE: ${{ github.workspace }}/${{ matrix.gemfile }}
     steps:
-      - uses: actions/checkout@v2
+      - uses: actions/checkout@v4
 
       - name: Install Ruby
         uses: ruby/setup-ruby@v1

Gemfile.lock

@@ -1,8 +1,9 @@
 PATH
   remote: .
   specs:
-    kamal (1.0.0)
+    kamal (1.4.0)
       activesupport (>= 7.0)
+      base64 (~> 0.2)
       bcrypt_pbkdf (~> 1.0)
       concurrent-ruby (~> 1.2)
       dotenv (~> 2.8)
@@ -15,82 +16,111 @@ PATH
 GEM
   remote: https://rubygems.org/
   specs:
-    actionpack (7.0.4.3)
-      actionview (= 7.0.4.3)
-      activesupport (= 7.0.4.3)
-      rack (~> 2.0, >= 2.2.0)
+    actionpack (7.1.2)
+      actionview (= 7.1.2)
+      activesupport (= 7.1.2)
+      nokogiri (>= 1.8.5)
+      racc
+      rack (>= 2.2.4)
+      rack-session (>= 1.0.1)
       rack-test (>= 0.6.3)
-      rails-dom-testing (~> 2.0)
-      rails-html-sanitizer (~> 1.0, >= 1.2.0)
-    actionview (7.0.4.3)
-      activesupport (= 7.0.4.3)
+      rails-dom-testing (~> 2.2)
+      rails-html-sanitizer (~> 1.6)
+    actionview (7.1.2)
+      activesupport (= 7.1.2)
       builder (~> 3.1)
-      erubi (~> 1.4)
-      rails-dom-testing (~> 2.0)
-      rails-html-sanitizer (~> 1.1, >= 1.2.0)
-    activesupport (7.0.4.3)
+      erubi (~> 1.11)
+      rails-dom-testing (~> 2.2)
+      rails-html-sanitizer (~> 1.6)
+    activesupport (7.1.2)
+      base64
+      bigdecimal
       concurrent-ruby (~> 1.0, >= 1.0.2)
+      connection_pool (>= 2.2.5)
+      drb
       i18n (>= 1.6, < 2)
       minitest (>= 5.1)
+      mutex_m
       tzinfo (~> 2.0)
+    base64 (0.2.0)
     bcrypt_pbkdf (1.1.0)
+    bigdecimal (3.1.5)
     builder (3.2.4)
     concurrent-ruby (1.2.2)
+    connection_pool (2.4.1)
     crass (1.0.6)
-    debug (1.7.2)
-      irb (>= 1.5.0)
-      reline (>= 0.3.1)
+    debug (1.9.1)
+      irb (~> 1.10)
+      reline (>= 0.3.8)
     dotenv (2.8.1)
+    drb (2.2.0)
+      ruby2_keywords
     ed25519 (1.3.0)
     erubi (1.12.0)
-    i18n (1.12.0)
+    i18n (1.14.1)
       concurrent-ruby (~> 1.0)
-    io-console (0.6.0)
-    irb (1.6.3)
-      reline (>= 0.3.0)
-    loofah (2.20.0)
+    io-console (0.7.1)
+    irb (1.11.0)
+      rdoc
+      reline (>= 0.3.8)
+    loofah (2.22.0)
       crass (~> 1.0.2)
-      nokogiri (>= 1.5.9)
-    method_source (1.0.0)
-    minitest (5.18.0)
-    mocha (2.0.2)
+      nokogiri (>= 1.12.0)
+    minitest (5.20.0)
+    mocha (2.1.0)
       ruby2_keywords (>= 0.0.5)
+    mutex_m (0.2.0)
     net-scp (4.0.0)
       net-ssh (>= 2.6.5, < 8.0.0)
-    net-ssh (7.1.0)
-    nokogiri (1.14.2-arm64-darwin)
+    net-ssh (7.2.1)
+    nokogiri (1.16.0-arm64-darwin)
       racc (~> 1.4)
-    nokogiri (1.14.2-x86_64-darwin)
+    nokogiri (1.16.0-x86_64-darwin)
       racc (~> 1.4)
-    nokogiri (1.14.2-x86_64-linux)
+    nokogiri (1.16.0-x86_64-linux)
       racc (~> 1.4)
-    racc (1.6.2)
-    rack (2.2.6.4)
+    psych (5.1.2)
+      stringio
+    racc (1.7.3)
+    rack (3.0.8)
+    rack-session (2.0.0)
+      rack (>= 3.0.0)
     rack-test (2.1.0)
       rack (>= 1.3)
-    rails-dom-testing (2.0.3)
-      activesupport (>= 4.2.0)
+    rackup (2.1.0)
+      rack (>= 3)
+      webrick (~> 1.8)
+    rails-dom-testing (2.2.0)
+      activesupport (>= 5.0.0)
+      minitest
       nokogiri (>= 1.6)
-    rails-html-sanitizer (1.5.0)
-      loofah (~> 2.19, >= 2.19.1)
-    railties (7.0.4.3)
-      actionpack (= 7.0.4.3)
-      activesupport (= 7.0.4.3)
-      method_source
+    rails-html-sanitizer (1.6.0)
+      loofah (~> 2.21)
+      nokogiri (~> 1.14)
+    railties (7.1.2)
+      actionpack (= 7.1.2)
+      activesupport (= 7.1.2)
+      irb
+      rackup (>= 1.0.0)
       rake (>= 12.2)
-      thor (~> 1.0)
-      zeitwerk (~> 2.5)
-    rake (13.0.6)
-    reline (0.3.3)
+      thor (~> 1.0, >= 1.2.2)
+      zeitwerk (~> 2.6)
+    rake (13.1.0)
+    rdoc (6.6.2)
+      psych (>= 4.0.0)
+    reline (0.4.2)
       io-console (~> 0.5)
     ruby2_keywords (0.0.5)
-    sshkit (1.21.4)
+    sshkit (1.21.7)
+      mutex_m
       net-scp (>= 1.1.2)
       net-ssh (>= 2.8.0)
-    thor (1.2.1)
+    stringio (3.1.0)
+    thor (1.3.0)
     tzinfo (2.0.6)
       concurrent-ruby (~> 1.0)
-    zeitwerk (2.6.7)
+    webrick (1.8.1)
+    zeitwerk (2.6.12)
 
 PLATFORMS
   arm64-darwin

gemfiles/ruby_2.7.gemfile

@@ -0,0 +1,6 @@
+source 'https://rubygems.org'
+git_source(:github) { |repo| "https://github.com/#{repo}.git" }
+
+gemspec path: "../"
+
+gem "nokogiri", "~> 1.15.0"

kamal.gemspec

@@ -20,6 +20,7 @@ Gem::Specification.new do |spec|
   spec.add_dependency "ed25519", "~> 1.2"
   spec.add_dependency "bcrypt_pbkdf", "~> 1.0"
   spec.add_dependency "concurrent-ruby", "~> 1.2"
+  spec.add_dependency "base64", "~> 0.2"
 
   spec.add_development_dependency "debug"
   spec.add_development_dependency "mocha"

lib/kamal/cli/accessory.rb

@@ -5,11 +5,11 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
       if name == "all"
         KAMAL.accessory_names.each { |accessory_name| boot(accessory_name) }
       else
-        with_accessory(name) do |accessory|
+        with_accessory(name) do |accessory, hosts|
           directories(name)
           upload(name)
 
-          on(accessory.hosts) do
+          on(hosts) do
             execute *KAMAL.registry.login if login
             execute *KAMAL.auditor.record("Booted #{name} accessory"), verbosity: :debug
             execute *accessory.run
@@ -22,8 +22,8 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
   desc "upload [NAME]", "Upload accessory files to host", hide: true
   def upload(name)
     mutating do
-      with_accessory(name) do |accessory|
-        on(accessory.hosts) do
+      with_accessory(name) do |accessory, hosts|
+        on(hosts) do
           accessory.files.each do |(local, remote)|
             accessory.ensure_local_file_present(local)
 
@@ -39,8 +39,8 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
   desc "directories [NAME]", "Create accessory directories on host", hide: true
   def directories(name)
     mutating do
-      with_accessory(name) do |accessory|
-        on(accessory.hosts) do
+      with_accessory(name) do |accessory, hosts|
+        on(hosts) do
           accessory.directories.keys.each do |host_path|
             execute *accessory.make_directory(host_path)
           end
@@ -49,11 +49,14 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
     end
   end
 
-  desc "reboot [NAME]", "Reboot existing accessory on host (stop container, remove container, start new container)"
+  desc "reboot [NAME]", "Reboot existing accessory on host (stop container, remove container, start new container; use NAME=all to boot all accessories)"
   def reboot(name)
     mutating do
-      with_accessory(name) do |accessory|
-        on(accessory.hosts) do
+      if name == "all"
+        KAMAL.accessory_names.each { |accessory_name| reboot(accessory_name) }
+      else
+        with_accessory(name) do |accessory, hosts|
+          on(hosts) do
             execute *KAMAL.registry.login
           end
 
@@ -63,12 +66,13 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
         end
       end
     end
+  end
 
   desc "start [NAME]", "Start existing accessory container on host"
   def start(name)
     mutating do
-      with_accessory(name) do |accessory|
-        on(accessory.hosts) do
+      with_accessory(name) do |accessory, hosts|
+        on(hosts) do
           execute *KAMAL.auditor.record("Started #{name} accessory"), verbosity: :debug
           execute *accessory.start
         end
@@ -79,8 +83,8 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
   desc "stop [NAME]", "Stop existing accessory container on host"
   def stop(name)
     mutating do
-      with_accessory(name) do |accessory|
-        on(accessory.hosts) do
+      with_accessory(name) do |accessory, hosts|
+        on(hosts) do
           execute *KAMAL.auditor.record("Stopped #{name} accessory"), verbosity: :debug
           execute *accessory.stop, raise_on_non_zero_exit: false
         end
@@ -103,8 +107,8 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
     if name == "all"
       KAMAL.accessory_names.each { |accessory_name| details(accessory_name) }
     else
-      with_accessory(name) do |accessory|
-        on(accessory.hosts) { puts capture_with_info(*accessory.info) }
+      with_accessory(name) do |accessory, hosts|
+        on(hosts) { puts capture_with_info(*accessory.info) }
       end
     end
   end
@@ -113,7 +117,7 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
   option :interactive, aliases: "-i", type: :boolean, default: false, desc: "Execute command over ssh for an interactive shell (use for console/bash)"
   option :reuse, type: :boolean, default: false, desc: "Reuse currently running container instead of starting a new one"
   def exec(name, cmd)
-    with_accessory(name) do |accessory|
+    with_accessory(name) do |accessory, hosts|
       case
       when options[:interactive] && options[:reuse]
         say "Launching interactive command with via SSH from existing container...", :magenta
@@ -125,14 +129,14 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
 
       when options[:reuse]
         say "Launching command from existing container...", :magenta
-        on(accessory.hosts) do
+        on(hosts) do
           execute *KAMAL.auditor.record("Executed cmd '#{cmd}' on #{name} accessory"), verbosity: :debug
           capture_with_info(*accessory.execute_in_existing_container(cmd))
         end
 
       else
         say "Launching command from new container...", :magenta
-        on(accessory.hosts) do
+        on(hosts) do
           execute *KAMAL.auditor.record("Executed cmd '#{cmd}' on #{name} accessory"), verbosity: :debug
           capture_with_info(*accessory.execute_in_new_container(cmd))
         end
@@ -146,12 +150,12 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
   option :grep, aliases: "-g", desc: "Show lines with grep match only (use this to fetch specific requests by id)"
   option :follow, aliases: "-f", desc: "Follow logs on primary server (or specific host set by --hosts)"
   def logs(name)
-    with_accessory(name) do |accessory|
+    with_accessory(name) do |accessory, hosts|
       grep = options[:grep]
 
       if options[:follow]
         run_locally do
-          info "Following logs on #{accessory.hosts}..."
+          info "Following logs on #{hosts}..."
           info accessory.follow_logs(grep: grep)
           exec accessory.follow_logs(grep: grep)
         end
@@ -159,7 +163,7 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
         since = options[:since]
         lines = options[:lines].presence || ((since || grep) ? nil : 100) # Default to 100 lines if since or grep isn't set
 
-        on(accessory.hosts) do
+        on(hosts) do
           puts capture_with_info(*accessory.logs(since: since, lines: lines, grep: grep))
         end
       end
@@ -188,8 +192,8 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
   desc "remove_container [NAME]", "Remove accessory container from host", hide: true
   def remove_container(name)
     mutating do
-      with_accessory(name) do |accessory|
-        on(accessory.hosts) do
+      with_accessory(name) do |accessory, hosts|
+        on(hosts) do
           execute *KAMAL.auditor.record("Remove #{name} accessory container"), verbosity: :debug
           execute *accessory.remove_container
         end
@@ -200,8 +204,8 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
   desc "remove_image [NAME]", "Remove accessory image from host", hide: true
   def remove_image(name)
     mutating do
-      with_accessory(name) do |accessory|
-        on(accessory.hosts) do
+      with_accessory(name) do |accessory, hosts|
+        on(hosts) do
           execute *KAMAL.auditor.record("Removed #{name} accessory image"), verbosity: :debug
           execute *accessory.remove_image
         end
@@ -212,8 +216,8 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
   desc "remove_service_directory [NAME]", "Remove accessory directory used for uploaded files and data directories from host", hide: true
   def remove_service_directory(name)
     mutating do
-      with_accessory(name) do |accessory|
-        on(accessory.hosts) do
+      with_accessory(name) do |accessory, hosts|
+        on(hosts) do
           execute *accessory.remove_service_directory
         end
       end
@@ -223,7 +227,7 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
   private
     def with_accessory(name)
       if accessory = KAMAL.accessory(name)
-        yield accessory
+        yield accessory, accessory_hosts(accessory)
       else
         error_on_missing_accessory(name)
       end
@@ -236,4 +240,12 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
         "No accessory by the name of '#{name}'" +
         (options ? " (options: #{options.to_sentence})" : "")
     end
+
+    def accessory_hosts(accessory)
+      if KAMAL.specific_hosts&.any?
+        KAMAL.specific_hosts & accessory.hosts
+      else
+        accessory.hosts
+      end
+    end
 end

lib/kamal/cli/app.rb

@@ -13,9 +13,8 @@ class Kamal::Cli::App < Kamal::Cli::Base
 
             KAMAL.roles_on(host).each do |role|
               app = KAMAL.app(role: role)
-              role_config = KAMAL.config.role(role)
 
-              if role_config.assets?
+              if role.assets?
                 execute *app.extract_assets
                 old_version = capture_with_info(*app.current_running_version, raise_on_non_zero_exit: false).strip
                 execute *app.sync_asset_volumes(old_version: old_version)
@@ -27,7 +26,6 @@ class Kamal::Cli::App < Kamal::Cli::Base
             KAMAL.roles_on(host).each do |role|
               app = KAMAL.app(role: role)
               auditor = KAMAL.auditor(role: role)
-              role_config = KAMAL.config.role(role)
 
               if capture_with_info(*app.container_id_for_version(version), raise_on_non_zero_exit: false).present?
                 tmp_version = "#{version}_replaced_#{SecureRandom.hex(8)}"
@@ -38,7 +36,7 @@ class Kamal::Cli::App < Kamal::Cli::Base
 
               old_version = capture_with_info(*app.current_running_version, raise_on_non_zero_exit: false).strip
 
-              execute *app.tie_cord(role_config.cord_host_file) if role_config.uses_cord?
+              execute *app.tie_cord(role.cord_host_file) if role.uses_cord?
 
               execute *auditor.record("Booted app version #{version}"), verbosity: :debug
 
@@ -47,7 +45,7 @@ class Kamal::Cli::App < Kamal::Cli::Base
               Kamal::Cli::Healthcheck::Poller.wait_for_healthy(pause_after_ready: true) { capture_with_info(*app.status(version: version)) }
 
               if old_version.present?
-                if role_config.uses_cord?
+                if role.uses_cord?
                   cord = capture_with_info(*app.cord(version: old_version), raise_on_non_zero_exit: false).strip
                   if cord.present?
                     execute *app.cut_cord(cord)
@@ -57,7 +55,7 @@ class Kamal::Cli::App < Kamal::Cli::Base
 
                 execute *app.stop(version: old_version), raise_on_non_zero_exit: false
 
-                execute *app.clean_up_assets if role_config.assets?
+                execute *app.clean_up_assets if role.assets?
               end
             end
           end
@@ -147,8 +145,12 @@ class Kamal::Cli::App < Kamal::Cli::Base
       using_version(version_or_latest) do |version|
         say "Launching command with version #{version} from new container...", :magenta
         on(KAMAL.hosts) do |host|
+          roles = KAMAL.roles_on(host)
+
+          roles.each do |role|
             execute *KAMAL.auditor.record("Executed cmd '#{cmd}' on app version #{version}"), verbosity: :debug
-          puts_by_host host, capture_with_info(*KAMAL.app.execute_in_new_container(cmd))
+            puts_by_host host, capture_with_info(*KAMAL.app(role: role).execute_in_new_container(cmd))
+          end
         end
       end
     end
@@ -198,19 +200,20 @@ class Kamal::Cli::App < Kamal::Cli::Base
     # FIXME: Catch when app containers aren't running
 
     grep = options[:grep]
-
+    since = options[:since]
     if options[:follow]
+      lines = options[:lines].presence || ((since || grep) ? nil : 10) # Default to 10 lines if since or grep isn't set
+
       run_locally do
         info "Following logs on #{KAMAL.primary_host}..."
 
         KAMAL.specific_roles ||= ["web"]
         role = KAMAL.roles_on(KAMAL.primary_host).first
 
-        info KAMAL.app(role: role).follow_logs(host: KAMAL.primary_host, grep: grep)
-        exec KAMAL.app(role: role).follow_logs(host: KAMAL.primary_host, grep: grep)
+        info KAMAL.app(role: role).follow_logs(host: KAMAL.primary_host, lines: lines, grep: grep)
+        exec KAMAL.app(role: role).follow_logs(host: KAMAL.primary_host, lines: lines, grep: grep)
       end
     else
-      since = options[:since]
       lines = options[:lines].presence || ((since || grep) ? nil : 100) # Default to 100 lines if since or grep isn't set
 
       on(KAMAL.hosts) do |host|

lib/kamal/cli/base.rb

@@ -14,8 +14,8 @@ module Kamal::Cli
     class_option :version, desc: "Run commands against a specific app version"
 
     class_option :primary, type: :boolean, aliases: "-p", desc: "Run commands only on primary host instead of all"
-    class_option :hosts, aliases: "-h", desc: "Run commands on these hosts instead of all (separate by comma)"
-    class_option :roles, aliases: "-r", desc: "Run commands on these roles instead of all (separate by comma)"
+    class_option :hosts, aliases: "-h", desc: "Run commands on these hosts instead of all (separate by comma, supports wildcards with *)"
+    class_option :roles, aliases: "-r", desc: "Run commands on these roles instead of all (separate by comma, supports wildcards with *)"
 
     class_option :config_file, aliases: "-c", default: "config/deploy.yml", desc: "Path to config file"
     class_option :destination, aliases: "-d", desc: "Specify destination to be used for config file (staging -> deploy.staging.yml)"
@@ -24,6 +24,7 @@ module Kamal::Cli
 
     def initialize(*)
       super
+      @original_env = ENV.to_h.dup
       load_envs
       initialize_commander(options_with_subcommand_class_options)
     end
@@ -37,6 +38,12 @@ module Kamal::Cli
         end
       end
 
+      def reload_envs
+        ENV.clear
+        ENV.update(@original_env)
+        load_envs
+      end
+
       def options_with_subcommand_class_options
         options.merge(@_initializer.last[:class_options] || {})
       end
@@ -75,8 +82,6 @@ module Kamal::Cli
       def mutating
         return yield if KAMAL.holding_lock?
 
-        KAMAL.config.ensure_env_available
-
         run_hook "pre-connect"
 
         ensure_run_directory
@@ -118,8 +123,9 @@ module Kamal::Cli
         yield
       rescue SSHKit::Runner::ExecuteError => e
         if e.message =~ /cannot create directory/
+          say "Deploy lock already in place!", :red
           on(KAMAL.primary_host) { puts capture_with_debug(*KAMAL.lock.status) }
-          raise LockError, "Deploy lock found"
+          raise LockError, "Deploy lock found. Run 'kamal lock help' for more information"
         else
           raise e
         end

lib/kamal/cli/env.rb

@@ -8,9 +8,8 @@ class Kamal::Cli::Env < Kamal::Cli::Base
         execute *KAMAL.auditor.record("Pushed env files"), verbosity: :debug
 
         KAMAL.roles_on(host).each do |role|
-          role_config = KAMAL.config.role(role)
           execute *KAMAL.app(role: role).make_env_directory
-          upload! StringIO.new(role_config.env_file), role_config.host_env_file_path, mode: 400
+          upload! StringIO.new(role.env_file), role.host_env_file_path, mode: 400
         end
       end
 
@@ -36,7 +35,6 @@ class Kamal::Cli::Env < Kamal::Cli::Base
         execute *KAMAL.auditor.record("Deleted env files"), verbosity: :debug
 
         KAMAL.roles_on(host).each do |role|
-          role_config = KAMAL.config.role(role)
           execute *KAMAL.app(role: role).remove_env_file
         end
       end

lib/kamal/cli/healthcheck.rb

@@ -3,6 +3,7 @@ class Kamal::Cli::Healthcheck < Kamal::Cli::Base
 
   desc "perform", "Health check current app version"
   def perform
+    raise "The primary host is not configured to run Traefik" unless KAMAL.config.role(KAMAL.config.primary_role).running_traefik?
     on(KAMAL.primary_host) do
       begin
         execute *KAMAL.healthcheck.run

lib/kamal/cli/main.rb

@@ -1,15 +1,18 @@
 class Kamal::Cli::Main < Kamal::Cli::Base
   desc "setup", "Setup all accessories, push the env, and deploy app to servers"
+  option :skip_push, aliases: "-P", type: :boolean, default: false, desc: "Skip image build and push"
   def setup
     print_runtime do
       mutating do
+        invoke_options = deploy_options
+
         say "Ensure Docker is installed...", :magenta
-        invoke "kamal:cli:server:bootstrap"
+        invoke "kamal:cli:server:bootstrap", [], invoke_options
 
         say "Push env files...", :magenta
-        invoke "kamal:cli:env:push"
+        invoke "kamal:cli:env:push", [], invoke_options
 
-        invoke "kamal:cli:accessory:boot", [ "all" ]
+        invoke "kamal:cli:accessory:boot", [ "all" ], invoke_options
         deploy
       end
     end
@@ -38,8 +41,10 @@ class Kamal::Cli::Main < Kamal::Cli::Base
         say "Ensure Traefik is running...", :magenta
         invoke "kamal:cli:traefik:boot", [], invoke_options
 
+        if KAMAL.config.role(KAMAL.config.primary_role).running_traefik?
           say "Ensure app can pass healthcheck...", :magenta
           invoke "kamal:cli:healthcheck:perform", [], invoke_options
+        end
 
         say "Detect stale containers...", :magenta
         invoke "kamal:cli:app:stale_containers", [], invoke_options.merge(stop: true)
@@ -170,6 +175,7 @@ class Kamal::Cli::Main < Kamal::Cli::Base
   end
 
   desc "envify", "Create .env by evaluating .env.erb (or .env.staging.erb -> .env.staging when using -d staging)"
+  option :skip_push, aliases: "-P", type: :boolean, default: false, desc: "Skip .env file push"
   def envify
     if destination = options[:destination]
       env_template_path = ".env.#{destination}.erb"
@@ -179,11 +185,13 @@ class Kamal::Cli::Main < Kamal::Cli::Base
       env_path          = ".env"
     end
 
-    File.write(env_path, ERB.new(File.read(env_template_path)).result, perm: 0600)
+    File.write(env_path, ERB.new(File.read(env_template_path), trim_mode: "-").result, perm: 0600)
 
-    load_envs # reload new file
+    unless options[:skip_push]
+      reload_envs
       invoke "kamal:cli:env:push", options
     end
+  end
 
   desc "remove", "Remove Traefik, app, accessories, and registry session from servers"
   option :confirmed, aliases: "-y", type: :boolean, default: false, desc: "Proceed without confirmation question"

lib/kamal/cli/prune.rb

@@ -18,12 +18,16 @@ class Kamal::Cli::Prune < Kamal::Cli::Base
     end
   end
 
-  desc "containers", "Prune all stopped containers, except the last 5"
+  desc "containers", "Prune all stopped containers, except the last n (default 5)"
+  option :retain, type: :numeric, default: nil, desc: "Number of containers to retain"
   def containers
+    retain = options.fetch(:retain, KAMAL.config.retain_containers)
+    raise "retain must be at least 1" if retain < 1
+
     mutating do
       on(KAMAL.hosts) do
         execute *KAMAL.auditor.record("Pruned containers"), verbosity: :debug
-        execute *KAMAL.prune.app_containers
+        execute *KAMAL.prune.app_containers(retain: retain)
         execute *KAMAL.prune.healthcheck_containers
       end
     end

lib/kamal/cli/server.rb

@@ -17,7 +17,9 @@ class Kamal::Cli::Server < Kamal::Cli::Base
     end
 
     if missing.any?
-      raise "Docker is not installed on #{missing.join(", ")} and can't be automatically installed without having root access and the `curl` command available. Install Docker manually: https://docs.docker.com/engine/install/"
-    end
+      raise "Docker is not installed on #{missing.join(", ")} and can't be automatically installed without having root access and either `wget` or `curl`. Install Docker manually: https://docs.docker.com/engine/install/"
+    end
+
+    run_hook "docker-setup"
   end
 end

lib/kamal/cli/templates/deploy.yml

@@ -77,9 +77,25 @@ registry:
 # Bridge fingerprinted assets, like JS and CSS, between versions to avoid
 # hitting 404 on in-flight requests. Combines all files from new and old
 # version inside the asset_path.
+#
+# If your app is using the Sprockets gem, ensure it sets `config.assets.manifest`.
+# See https://github.com/basecamp/kamal/issues/626 for details
+#
 # asset_path: /rails/public/assets
 
 # Configure rolling deploys by setting a wait time between batches of restarts.
 # boot:
 #   limit: 10 # Can also specify as a percentage of total hosts, such as "25%"
 #   wait: 2
+
+# Configure the role used to determine the primary_host. This host takes
+# deploy locks, runs health checks during the deploy, and follow logs, etc.
+#
+# Caution: there's no support for role renaming yet, so be careful to cleanup
+#          the previous role on the deployed hosts.
+# primary_role: web
+
+# Controls if we abort when see a role with no hosts. Disabling this may be
+# useful for more complex deploy configurations.
+#
+# allow_empty_roles: false

lib/kamal/cli/templates/sample_hooks/docker-setup.sample

@@ -0,0 +1,7 @@
+#!/usr/bin/env ruby
+
+# A sample docker-setup hook
+#
+# Sets up a Docker network which can then be used by the application’s containers
+
+ssh user@example.com docker network create kamal

lib/kamal/cli/templates/sample_hooks/post-traefik-reboot.sample

@@ -0,0 +1,3 @@
+#!/bin/sh
+
+echo "Rebooted Traefik on $KAMAL_HOSTS"

lib/kamal/cli/templates/sample_hooks/pre-build.sample

@@ -32,7 +32,7 @@ fi
 current_branch=$(git branch --show-current)
 
 if [ -z "$current_branch" ]; then
-  echo "No git remote set, aborting..." >&2
+  echo "Not on a git branch, aborting..." >&2
   exit 1
 fi
 

lib/kamal/cli/templates/sample_hooks/pre-traefik-reboot.sample

@@ -0,0 +1,3 @@
+#!/bin/sh
+
+echo "Rebooting Traefik on $KAMAL_HOSTS..."

lib/kamal/cli/traefik.rb

@@ -13,13 +13,19 @@ class Kamal::Cli::Traefik < Kamal::Cli::Base
   option :rolling, type: :boolean, default: false, desc: "Reboot traefik on hosts in sequence, rather than in parallel"
   def reboot
     mutating do
-      on(KAMAL.traefik_hosts, in: options[:rolling] ? :sequence : :parallel) do
+      host_groups = options[:rolling] ? KAMAL.traefik_hosts : [KAMAL.traefik_hosts]
+      host_groups.each do |hosts|
+        host_list = Array(hosts).join(",")
+        run_hook "pre-traefik-reboot", hosts: host_list
+        on(hosts) do
           execute *KAMAL.auditor.record("Rebooted traefik"), verbosity: :debug
           execute *KAMAL.registry.login
-        execute *KAMAL.traefik.stop
+          execute *KAMAL.traefik.stop, raise_on_non_zero_exit: false
           execute *KAMAL.traefik.remove_container
           execute *KAMAL.traefik.run
         end
+        run_hook "post-traefik-reboot", hosts: host_list
+      end
     end
   end
 
@@ -38,7 +44,7 @@ class Kamal::Cli::Traefik < Kamal::Cli::Base
     mutating do
       on(KAMAL.traefik_hosts) do
         execute *KAMAL.auditor.record("Stopped traefik"), verbosity: :debug
-        execute *KAMAL.traefik.stop
+        execute *KAMAL.traefik.stop, raise_on_non_zero_exit: false
       end
     end
   end

lib/kamal/commander.rb

@@ -24,19 +24,36 @@ class Kamal::Commander
   attr_reader :specific_roles, :specific_hosts
 
   def specific_primary!
-    self.specific_hosts = [ config.primary_web_host ]
+    self.specific_hosts = [ config.primary_host ]
   end
 
   def specific_roles=(role_names)
-    @specific_roles = config.roles.select { |r| role_names.include?(r.name) } if role_names.present?
+    if role_names.present?
+      @specific_roles = Kamal::Utils.filter_specific_items(role_names, config.roles)
+
+      if @specific_roles.empty?
+        raise ArgumentError, "No --roles match for #{role_names.join(',')}"
+      end
+
+      @specific_roles
+    end
   end
 
   def specific_hosts=(hosts)
-    @specific_hosts = config.all_hosts & hosts if hosts.present?
+    if hosts.present?
+      @specific_hosts = Kamal::Utils.filter_specific_items(hosts, config.all_hosts)
+
+      if @specific_hosts.empty?
+        raise ArgumentError, "No --hosts match for #{hosts.join(',')}"
+      end
+
+      @specific_hosts
+    end
   end
 
   def primary_host
-    specific_hosts&.first || specific_roles&.first&.primary_host || config.primary_web_host
+    # Given a list of specific roles, make an effort to match up with the primary_role
+    specific_hosts&.first || specific_roles&.detect { |role| role == config.primary_role }&.primary_host || specific_roles&.first&.primary_host || config.primary_host
   end
 
   def primary_role
@@ -56,7 +73,7 @@ class Kamal::Commander
   end
 
   def roles_on(host)
-    roles.select { |role| role.hosts.include?(host.to_s) }.map(&:name)
+    roles.select { |role| role.hosts.include?(host.to_s) }
   end
 
   def traefik_hosts

lib/kamal/commands/app.rb

@@ -3,12 +3,11 @@ class Kamal::Commands::App < Kamal::Commands::Base
 
   ACTIVE_DOCKER_STATUSES = [ :running, :restarting ]
 
-  attr_reader :role, :role_config
+  attr_reader :role, :role
 
   def initialize(config, role: nil)
     super(config)
     @role = role
-    @role_config = config.role(self.role)
   end
 
   def run(hostname: nil)
@@ -18,15 +17,16 @@ class Kamal::Commands::App < Kamal::Commands::Base
       "--name", container_name,
       *(["--hostname", hostname] if hostname),
       "-e", "KAMAL_CONTAINER_NAME=\"#{container_name}\"",
-      *role_config.env_args,
-      *role_config.health_check_args,
-      *config.logging_args,
+      "-e", "KAMAL_VERSION=\"#{config.version}\"",
+      *role.env_args,
+      *role.health_check_args,
+      *role.logging_args,
       *config.volume_args,
-      *role_config.asset_volume_args,
-      *role_config.label_args,
-      *role_config.option_args,
+      *role.asset_volume_args,
+      *role.label_args,
+      *role.option_args,
       config.absolute_image,
-      role_config.cmd
+      role.cmd
   end
 
   def start
@@ -63,22 +63,22 @@ class Kamal::Commands::App < Kamal::Commands::Base
   def list_versions(*docker_args, statuses: nil)
     pipe \
       docker(:ps, *filter_args(statuses: statuses), *docker_args, "--format", '"{{.Names}}"'),
-      %(while read line; do echo ${line##{role_config.container_prefix}-}; done) # Extract SHA from "service-role-dest-SHA"
+      %(while read line; do echo ${line##{role.container_prefix}-}; done) # Extract SHA from "service-role-dest-SHA"
   end
 
 
   def make_env_directory
-    make_directory role_config.host_env_directory
+    make_directory role.host_env_directory
   end
 
   def remove_env_file
-    [ :rm, "-f", role_config.host_env_file_path ]
+    [ :rm, "-f", role.host_env_file_path ]
   end
 
 
   private
     def container_name(version = nil)
-      [ role_config.container_prefix, version || config.version ].compact.join("-")
+      [ role.container_prefix, version || config.version ].compact.join("-")
     end
 
     def filter_args(statuses: nil)

lib/kamal/commands/app/assets.rb

@@ -1,20 +1,20 @@
 module Kamal::Commands::App::Assets
   def extract_assets
-    asset_container = "#{role_config.container_prefix}-assets"
+    asset_container = "#{role.container_prefix}-assets"
 
     combine \
-      make_directory(role_config.asset_extracted_path),
+      make_directory(role.asset_extracted_path),
       [*docker(:stop, "-t 1", asset_container, "2> /dev/null"), "|| true"],
       docker(:run, "--name", asset_container, "--detach", "--rm", config.latest_image, "sleep 1000000"),
-      docker(:cp, "-L", "#{asset_container}:#{role_config.asset_path}/.", role_config.asset_extracted_path),
+      docker(:cp, "-L", "#{asset_container}:#{role.asset_path}/.", role.asset_extracted_path),
       docker(:stop, "-t 1", asset_container),
       by: "&&"
   end
 
   def sync_asset_volumes(old_version: nil)
-    new_extracted_path, new_volume_path = role_config.asset_extracted_path(config.version), role_config.asset_volume.host_path
+    new_extracted_path, new_volume_path = role.asset_extracted_path(config.version), role.asset_volume.host_path
     if old_version.present?
-      old_extracted_path, old_volume_path = role_config.asset_extracted_path(old_version), role_config.asset_volume(old_version).host_path
+      old_extracted_path, old_volume_path = role.asset_extracted_path(old_version), role.asset_volume(old_version).host_path
     end
 
     commands = [make_directory(new_volume_path), copy_contents(new_extracted_path, new_volume_path)]
@@ -29,8 +29,8 @@ module Kamal::Commands::App::Assets
 
   def clean_up_assets
     chain \
-      find_and_remove_older_siblings(role_config.asset_extracted_path),
-      find_and_remove_older_siblings(role_config.asset_volume_path)
+      find_and_remove_older_siblings(role.asset_extracted_path),
+      find_and_remove_older_siblings(role.asset_volume_path)
   end
 
   private
@@ -39,7 +39,7 @@ module Kamal::Commands::App::Assets
         :find,
         Pathname.new(path).dirname.to_s,
         "-maxdepth 1",
-        "-name", "'#{role_config.container_prefix}-*'",
+        "-name", "'#{role.container_prefix}-*'",
         "!", "-name", Pathname.new(path).basename.to_s,
         "-exec rm -rf \"{}\" +"
       ]

lib/kamal/commands/app/cord.rb

@@ -2,7 +2,7 @@ module Kamal::Commands::App::Cord
   def cord(version:)
     pipe \
       docker(:inspect, "-f '{{ range .Mounts }}{{printf \"%s %s\\n\" .Source .Destination}}{{ end }}'", container_name(version)),
-      [:awk, "'$2 == \"#{role_config.cord_volume.container_path}\" {print $1}'"]
+      [:awk, "'$2 == \"#{role.cord_volume.container_path}\" {print $1}'"]
   end
 
   def tie_cord(cord)

lib/kamal/commands/app/execution.rb

@@ -10,9 +10,9 @@ module Kamal::Commands::App::Execution
     docker :run,
       ("-it" if interactive),
       "--rm",
-      *role_config&.env_args,
+      *role&.env_args,
       *config.volume_args,
-      *role_config&.option_args,
+      *role&.option_args,
       config.absolute_image,
       *command
   end

lib/kamal/commands/app/logging.rb

@@ -6,11 +6,11 @@ module Kamal::Commands::App::Logging
       ("grep '#{grep}'" if grep)
   end
 
-  def follow_logs(host:, grep: nil)
+  def follow_logs(host:, lines: nil, grep: nil)
     run_over_ssh \
       pipe(
         current_running_container_id,
-        "xargs docker logs --timestamps --tail 10 --follow 2>&1",
+        "xargs docker logs --timestamps#{" --tail #{lines}" if lines} --follow 2>&1",
         (%(grep "#{grep}") if grep)
       ),
       host: host

lib/kamal/commands/base.rb

@@ -18,7 +18,7 @@ module Kamal::Commands
         elsif config.ssh.proxy && config.ssh.proxy.is_a?(Net::SSH::Proxy::Command)
           cmd << " -o ProxyCommand='#{config.ssh.proxy.command_line_template}'"
         end
-        cmd << " -t #{config.ssh.user}@#{host} '#{command.join(" ")}'"
+        cmd << " -t #{config.ssh.user}@#{host} -p #{config.ssh.port} '#{command.join(" ")}'"
       end
     end
 
@@ -62,10 +62,18 @@ module Kamal::Commands
         combine *commands, by: ">"
       end
 
+      def any(*commands)
+        combine *commands, by: "||"
+      end
+
       def xargs(command)
         [ :xargs, command ].flatten
       end
 
+      def shell(command)
+        [ :sh, "-c", "'#{command.flatten.join(" ").gsub("'", "'\\''")}'" ]
+      end
+
       def docker(*args)
         args.compact.unshift :docker
       end

lib/kamal/commands/builder/base.rb

@@ -3,7 +3,7 @@ class Kamal::Commands::Builder::Base < Kamal::Commands::Base
   class BuilderError < StandardError; end
 
   delegate :argumentize, to: Kamal::Utils
-  delegate :args, :secrets, :dockerfile, :local_arch, :local_host, :remote_arch, :remote_host, :cache_from, :cache_to, to: :builder_config
+  delegate :args, :secrets, :dockerfile, :local_arch, :local_host, :remote_arch, :remote_host, :cache_from, :cache_to, :ssh, to: :builder_config
 
   def clean
     docker :image, :rm, "--force", config.absolute_image
@@ -14,7 +14,7 @@ class Kamal::Commands::Builder::Base < Kamal::Commands::Base
   end
 
   def build_options
-    [ *build_tags, *build_cache, *build_labels, *build_args, *build_secrets, *build_dockerfile ]
+    [ *build_tags, *build_cache, *build_labels, *build_args, *build_secrets, *build_dockerfile, *build_ssh ]
   end
 
   def build_context
@@ -24,7 +24,10 @@ class Kamal::Commands::Builder::Base < Kamal::Commands::Base
   def validate_image
     pipe \
       docker(:inspect, "-f", "'{{ .Config.Labels.service }}'", config.absolute_image),
-      [:grep, "-x", config.service, "||", "(echo \"Image #{config.absolute_image} is missing the `service` label\" && exit 1)"]
+      any(
+        [:grep, "-x", config.service],
+        "(echo \"Image #{config.absolute_image} is missing the 'service' label\" && exit 1)"
+      )
   end
 
 
@@ -60,6 +63,10 @@ class Kamal::Commands::Builder::Base < Kamal::Commands::Base
       end
     end
 
+    def build_ssh
+      argumentize "--ssh", ssh if ssh.present?
+    end
+
     def builder_config
       config.builder
     end

lib/kamal/commands/builder/multiarch.rb

@@ -10,7 +10,7 @@ class Kamal::Commands::Builder::Multiarch < Kamal::Commands::Builder::Base
   def push
     docker :buildx, :build,
       "--push",
-      "--platform", "linux/amd64,linux/arm64",
+      "--platform", platform_names,
       "--builder", builder_name,
       *build_options,
       build_context
@@ -26,4 +26,12 @@ class Kamal::Commands::Builder::Multiarch < Kamal::Commands::Builder::Base
     def builder_name
       "kamal-#{config.service}-multiarch"
     end
+
+    def platform_names
+      if local_arch
+        "linux/#{local_arch}"
+      else
+        "linux/amd64,linux/arm64"
+      end
+    end
 end

lib/kamal/commands/docker.rb

@@ -1,7 +1,7 @@
 class Kamal::Commands::Docker < Kamal::Commands::Base
   # Install Docker using the https://github.com/docker/docker-install convenience script.
   def install
-    pipe [ :curl, "-fsSL", "https://get.docker.com" ], :sh
+    pipe get_docker, :sh
   end
 
   # Checks the Docker client version. Fails if Docker is not installed.
@@ -16,6 +16,15 @@ class Kamal::Commands::Docker < Kamal::Commands::Base
 
   # Do we have superuser access to install Docker and start system services?
   def superuser?
-    [ '[ "${EUID:-$(id -u)}" -eq 0 ]' ]
+    [ '[ "${EUID:-$(id -u)}" -eq 0 ] || command -v sudo >/dev/null || command -v su >/dev/null' ]
+  end
+
+  private
+    def get_docker
+      shell \
+        any \
+          [ :curl, "-fsSL", "https://get.docker.com" ],
+          [ :wget, "-O -", "https://get.docker.com" ],
+          [ :echo, "\"exit 1\"" ]
     end
 end

lib/kamal/commands/healthcheck.rb

@@ -1,7 +1,7 @@
 class Kamal::Commands::Healthcheck < Kamal::Commands::Base
 
   def run
-    web = config.role(:web)
+    primary = config.role(config.primary_role)
 
     docker :run,
       "--detach",
@@ -9,12 +9,12 @@ class Kamal::Commands::Healthcheck < Kamal::Commands::Base
       "--publish", "#{exposed_port}:#{config.healthcheck["port"]}",
       "--label", "service=#{config.healthcheck_service}",
       "-e", "KAMAL_CONTAINER_NAME=\"#{config.healthcheck_service}\"",
-      *web.env_args,
-      *web.health_check_args(cord: false),
+      *primary.env_args,
+      *primary.health_check_args(cord: false),
       *config.volume_args,
-      *web.option_args,
+      *primary.option_args,
       config.absolute_image,
-      web.cmd
+      primary.cmd
   end
 
   def status

lib/kamal/commands/lock.rb

@@ -1,5 +1,6 @@
 require "active_support/duration"
 require "time"
+require "base64"
 
 class Kamal::Commands::Lock < Kamal::Commands::Base
   def acquire(message, version)

lib/kamal/commands/prune.rb

@@ -13,10 +13,10 @@ class Kamal::Commands::Prune < Kamal::Commands::Base
       "while read image tag; do docker rmi $tag; done"
   end
 
-  def app_containers(keep_last: 5)
+  def app_containers(retain:)
     pipe \
       docker(:ps, "-q", "-a", *service_filter, *stopped_containers_filters),
-      "tail -n +#{keep_last + 1}",
+      "tail -n +#{retain + 1}",
       "while read container_id; do docker rm $container_id; done"
   end
 

lib/kamal/commands/registry.rb

@@ -2,7 +2,10 @@ class Kamal::Commands::Registry < Kamal::Commands::Base
   delegate :registry, to: :config
 
   def login
-    docker :login, registry["server"], "-u", sensitive(lookup("username")), "-p", sensitive(lookup("password"))
+    docker :login,
+      registry["server"],
+      "-u", sensitive(Kamal::Utils.escape_shell_value(lookup("username"))),
+      "-p", sensitive(Kamal::Utils.escape_shell_value(lookup("password")))
   end
 
   def logout

lib/kamal/commands/traefik.rb

@@ -1,11 +1,19 @@
 class Kamal::Commands::Traefik < Kamal::Commands::Base
   delegate :argumentize, :optionize, to: Kamal::Utils
 
-  DEFAULT_IMAGE = "traefik:v2.9"
+  DEFAULT_IMAGE = "traefik:v2.10"
   CONTAINER_PORT = 80
   DEFAULT_ARGS = {
     'log.level' => 'DEBUG'
   }
+  DEFAULT_LABELS = {
+    # These ensure we serve a 502 rather than a 404 if no containers are available
+    "traefik.http.routers.catchall.entryPoints" => "http",
+    "traefik.http.routers.catchall.rule" => "PathPrefix(`/`)",
+    "traefik.http.routers.catchall.service" => "unavailable",
+    "traefik.http.routers.catchall.priority" => 1,
+    "traefik.http.services.unavailable.loadbalancer.server.port" => "0"
+  }
 
   def run
     docker :run, "--name traefik",
@@ -31,7 +39,7 @@ class Kamal::Commands::Traefik < Kamal::Commands::Base
   end
 
   def start_or_run
-    combine start, run, by: "||"
+    any start, run
   end
 
   def info
@@ -97,7 +105,7 @@ class Kamal::Commands::Traefik < Kamal::Commands::Base
     end
 
     def labels
-      config.traefik["labels"] || []
+      DEFAULT_LABELS.merge(config.traefik["labels"] || {})
     end
 
     def image

lib/kamal/configuration.rb

@@ -6,7 +6,7 @@ require "erb"
 require "net/ssh/proxy/jump"
 
 class Kamal::Configuration
-  delegate :service, :image, :servers, :env, :labels, :registry, :stop_wait_time, :hooks_path, to: :raw_config, allow_nil: true
+  delegate :service, :image, :servers, :env, :labels, :registry, :stop_wait_time, :hooks_path, :logging, to: :raw_config, allow_nil: true
   delegate :argumentize, :optionize, to: Kamal::Utils
 
   attr_reader :destination, :raw_config
@@ -25,7 +25,9 @@ class Kamal::Configuration
 
       def load_config_file(file)
         if file.exist?
-          YAML.load(ERB.new(IO.read(file)).result).symbolize_keys
+          # Newer Psych doesn't load aliases by default
+          load_method = YAML.respond_to?(:unsafe_load) ? :unsafe_load : :load
+          YAML.send(load_method, ERB.new(IO.read(file)).result).symbolize_keys
         else
           raise "Configuration file not found in #{file}"
         end
@@ -89,15 +91,34 @@ class Kamal::Configuration
     roles.flat_map(&:hosts).uniq
   end
 
-  def primary_web_host
-    role(:web).primary_host
+  def primary_host
+    primary_role&.primary_host
+  end
+
+  def primary_role_name
+    raw_config.primary_role || "web"
+  end
+
+  def primary_role
+    role(primary_role_name)
+  end
+
+  def allow_empty_roles?
+    raw_config.allow_empty_roles
+  end
+
+  def traefik_roles
+    roles.select(&:running_traefik?)
+  end
+
+  def traefik_role_names
+    traefik_roles.flat_map(&:name)
   end
 
   def traefik_hosts
-    roles.select(&:running_traefik?).flat_map(&:hosts).uniq
+    traefik_roles.flat_map(&:hosts).uniq
   end
 
-
   def repository
     [ raw_config.registry["server"], image ].compact.join("/")
   end
@@ -118,6 +139,10 @@ class Kamal::Configuration
     raw_config.require_destination
   end
 
+  def retain_containers
+    raw_config.retain_containers || 5
+  end
+
 
   def volume_args
     if raw_config.volumes.present?
@@ -128,9 +153,9 @@ class Kamal::Configuration
   end
 
   def logging_args
-    if raw_config.logging.present?
-      optionize({ "log-driver" => raw_config.logging["driver"] }.compact) +
-        argumentize("--log-opt", raw_config.logging["options"])
+    if logging.present?
+      optionize({ "log-driver" => logging["driver"] }.compact) +
+        argumentize("--log-opt", logging["options"])
     else
       argumentize("--log-opt", { "max-size" => "10m" })
     end
@@ -201,21 +226,14 @@ class Kamal::Configuration
 
 
   def valid?
-    ensure_destination_if_required && ensure_required_keys_present && ensure_valid_kamal_version
-  end
-
-  # Will raise KeyError if any secret ENVs are missing
-  def ensure_env_available
-    roles.collect(&:env_file).each(&:to_s)
-
-    true
+    ensure_destination_if_required && ensure_required_keys_present && ensure_valid_kamal_version && ensure_retain_containers_valid && ensure_valid_service_name
   end
 
   def to_h
     {
       roles: role_names,
       hosts: all_hosts,
-      primary_host: primary_web_host,
+      primary_host: primary_host,
       version: version,
       repository: repository,
       absolute_image: absolute_image,
@@ -254,11 +272,27 @@ class Kamal::Configuration
         raise ArgumentError, "You must specify a password for the registry in config/deploy.yml (or set the ENV variable if that's used)"
       end
 
+      unless role_names.include?(primary_role_name)
+        raise ArgumentError, "The primary_role #{primary_role_name} isn't defined"
+      end
+
+      if primary_role.hosts.empty?
+        raise ArgumentError, "No servers specified for the #{primary_role.name} primary_role"
+      end
+
+      unless allow_empty_roles?
         roles.each do |role|
           if role.hosts.empty?
-          raise ArgumentError, "No servers specified for the #{role.name} role"
+            raise ArgumentError, "No servers specified for the #{role.name} role. You can ignore this with allow_empty_roles: true"
           end
         end
+      end
+
+      true
+    end
+
+    def ensure_valid_service_name
+      raise ArgumentError, "Service name can only include alphanumeric characters, hyphens, and underscores" unless raw_config[:service] =~ /^[a-z0-9_-]+$/
 
       true
     end
@@ -271,6 +305,12 @@ class Kamal::Configuration
       true
     end
 
+    def ensure_retain_containers_valid
+      raise ArgumentError, "Must retain at least 1 container" if retain_containers < 1
+
+      true
+    end
+
 
     def role_names
       raw_config.servers.is_a?(Array) ? [ "web" ] : raw_config.servers.keys.sort

lib/kamal/configuration/accessory.rb

@@ -8,7 +8,7 @@ class Kamal::Configuration::Accessory
   end
 
   def service_name
-    "#{config.service}-#{name}"
+    specifics["service"] || "#{config.service}-#{name}"
   end
 
   def image
@@ -70,8 +70,8 @@ class Kamal::Configuration::Accessory
 
   def directories
     specifics["directories"]&.to_h do |host_to_container_mapping|
-      host_relative_path, container_path = host_to_container_mapping.split(":")
-      [ expand_host_path(host_relative_path), container_path ]
+      host_path, container_path = host_to_container_mapping.split(":")
+      [ expand_host_path(host_path), container_path ]
     end || {}
   end
 
@@ -138,13 +138,17 @@ class Kamal::Configuration::Accessory
 
     def remote_directories_as_volumes
       specifics["directories"]&.collect do |host_to_container_mapping|
-        host_relative_path, container_path = host_to_container_mapping.split(":")
-        [ expand_host_path(host_relative_path), container_path ].join(":")
+        host_path, container_path = host_to_container_mapping.split(":")
+        [ expand_host_path(host_path), container_path ].join(":")
       end || []
     end
 
-    def expand_host_path(host_relative_path)
-      "#{service_data_directory}/#{host_relative_path}"
+    def expand_host_path(host_path)
+      absolute_path?(host_path) ? host_path : "#{service_data_directory}/#{host_path}"
+    end
+
+    def absolute_path?(path)
+      Pathname.new(path).absolute?
     end
 
     def service_data_directory

lib/kamal/configuration/boot.rb

@@ -8,7 +8,7 @@ class Kamal::Configuration::Boot
     limit = @options["limit"]
 
     if limit.to_s.end_with?("%")
-      @host_count * limit.to_i / 100
+      [@host_count * limit.to_i / 100, 1].max
     else
       limit
     end

lib/kamal/configuration/builder.rb

@@ -81,6 +81,10 @@ class Kamal::Configuration::Builder
     end
   end
 
+  def ssh
+    @options["ssh"]
+  end
+
   private
     def valid?
       if @options["cache"] && @options["cache"]["type"]

lib/kamal/configuration/role.rb

@@ -3,6 +3,7 @@ class Kamal::Configuration::Role
   delegate :argumentize, :optionize, to: Kamal::Utils
 
   attr_accessor :name
+  alias to_s name
 
   def initialize(name, config:)
     @name, @config = name.inquiry, config
@@ -36,6 +37,18 @@ class Kamal::Configuration::Role
     argumentize "--label", labels
   end
 
+  def logging_args
+    args = config.logging || {}
+    args.deep_merge!(specializations["logging"]) if specializations["logging"].present?
+
+    if args.any?
+      optionize({ "log-driver" => args["driver"] }.compact) +
+        argumentize("--log-opt", args["options"])
+    else
+      config.logging_args
+    end
+  end
+
 
   def env
     if config.env && config.env["secret"]
@@ -93,7 +106,15 @@ class Kamal::Configuration::Role
 
 
   def running_traefik?
-    name.web? || specializations["traefik"]
+    if specializations["traefik"].nil?
+      primary?
+    else
+      specializations["traefik"]
+    end
+  end
+
+  def primary?
+    self == @config.primary_role
   end
 
 
@@ -185,6 +206,7 @@ class Kamal::Configuration::Role
           "traefik.http.services.#{traefik_service}.loadbalancer.server.scheme" => "http",
 
           "traefik.http.routers.#{traefik_service}.rule" => "PathPrefix(`/`)",
+          "traefik.http.routers.#{traefik_service}.priority" => "2",
           "traefik.http.middlewares.#{traefik_service}-retry.retry.attempts" => "5",
           "traefik.http.middlewares.#{traefik_service}-retry.retry.initialinterval" => "500ms",
           "traefik.http.routers.#{traefik_service}.middlewares" => "#{traefik_service}-retry@docker"
@@ -230,7 +252,7 @@ class Kamal::Configuration::Role
         clear_app_env  = config.env["secret"] ? Array(config.env["clear"]) : Array(config.env["clear"] || config.env)
         clear_role_env = specialized_env["secret"] ? Array(specialized_env["clear"]) : Array(specialized_env["clear"] || specialized_env)
 
-        new_env["clear"] = (clear_app_env + clear_role_env).uniq
+        new_env["clear"] = clear_app_env.to_h.merge(clear_role_env.to_h)
       end
     end
 

lib/kamal/configuration/ssh.rb

@@ -9,6 +9,10 @@ class Kamal::Configuration::Ssh
     config.fetch("user", "root")
   end
 
+  def port
+    config.fetch("port", 22)
+  end
+
   def proxy
     if (proxy = config["proxy"])
       Net::SSH::Proxy::Jump.new(proxy.include?("@") ? proxy : "root@#{proxy}")
@@ -18,7 +22,7 @@ class Kamal::Configuration::Ssh
   end
 
   def options
-    { user: user, proxy: proxy, logger: logger, keepalive: true, keepalive_interval: 30 }.compact
+    { user: user, port: port, proxy: proxy, logger: logger, keepalive: true, keepalive_interval: 30 }.compact
   end
 
   def to_h

lib/kamal/sshkit_with_ext.rb

@@ -1,5 +1,6 @@
 require "sshkit"
 require "sshkit/dsl"
+require "net/scp"
 require "active_support/core_ext/hash/deep_merge"
 require "json"
 

lib/kamal/utils.rb

@@ -58,4 +58,20 @@ module Kamal::Utils
       .gsub(/`/, '\\\\`')
       .gsub(DOLLAR_SIGN_WITHOUT_SHELL_EXPANSION_REGEX, '\$')
   end
+
+  # Apply a list of host or role filters, including wildcard matches
+  def filter_specific_items(filters, items)
+    matches = []
+
+    Array(filters).select do |filter|
+      matches += Array(items).select do |item|
+        # Only allow * for a wildcard
+        pattern = Regexp.escape(filter).gsub('\*', '.*')
+        # items are roles or hosts
+        (item.respond_to?(:name) ? item.name : item).match(/^#{pattern}$/)
+      end
+    end
+
+    matches
+  end
 end

lib/kamal/utils/sensitive.rb

@@ -1,4 +1,5 @@
 require "active_support/core_ext/module/delegation"
+require "sshkit"
 
 class Kamal::Utils::Sensitive
   # So SSHKit knows to redact these values.

lib/kamal/version.rb

@@ -1,3 +1,3 @@
 module Kamal
-  VERSION = "1.0.0"
+  VERSION = "1.4.0"
 end

test/cli/accessory_test.rb

@@ -48,6 +48,18 @@ class CliAccessoryTest < CliTestCase
     run_command("reboot", "mysql")
   end
 
+  test "reboot all" do
+    Kamal::Commands::Registry.any_instance.expects(:login).times(3)
+    Kamal::Cli::Accessory.any_instance.expects(:stop).with("mysql")
+    Kamal::Cli::Accessory.any_instance.expects(:remove_container).with("mysql")
+    Kamal::Cli::Accessory.any_instance.expects(:boot).with("mysql", login: false)
+    Kamal::Cli::Accessory.any_instance.expects(:stop).with("redis")
+    Kamal::Cli::Accessory.any_instance.expects(:remove_container).with("redis")
+    Kamal::Cli::Accessory.any_instance.expects(:boot).with("redis", login: false)
+
+    run_command("reboot", "all")
+  end
+
   test "start" do
     assert_match "docker container start app-mysql", run_command("start", "mysql")
   end
@@ -97,7 +109,7 @@ class CliAccessoryTest < CliTestCase
 
   test "logs with follow" do
     SSHKit::Backend::Abstract.any_instance.stubs(:exec)
-      .with("ssh -t root@1.1.1.3 'docker logs app-mysql --timestamps --tail 10 --follow 2>&1'")
+      .with("ssh -t root@1.1.1.3 -p 22 'docker logs app-mysql --timestamps --tail 10 --follow 2>&1'")
 
     assert_match "docker logs app-mysql --timestamps --tail 10 --follow 2>&1", run_command("logs", "mysql", "--follow")
   end
@@ -136,6 +148,30 @@ class CliAccessoryTest < CliTestCase
     assert_match "rm -rf app-mysql", run_command("remove_service_directory", "mysql")
   end
 
+  test "hosts param respected" do
+    Kamal::Cli::Accessory.any_instance.expects(:directories).with("redis")
+    Kamal::Cli::Accessory.any_instance.expects(:upload).with("redis")
+
+    run_command("boot", "redis", "--hosts", "1.1.1.1").tap do |output|
+      assert_match /docker login.*on 1.1.1.1/, output
+      refute_match /docker login.*on 1.1.1.2/, output
+      assert_match "docker run --name app-redis --detach --restart unless-stopped --log-opt max-size=\"10m\" --publish 6379:6379 --env-file .kamal/env/accessories/app-redis.env --volume $PWD/app-redis/data:/data --label service=\"app-redis\" redis:latest on 1.1.1.1", output
+      refute_match "docker run --name app-redis --detach --restart unless-stopped --log-opt max-size=\"10m\" --publish 6379:6379 --env-file .kamal/env/accessories/app-redis.env --volume $PWD/app-redis/data:/data --label service=\"app-redis\" redis:latest on 1.1.1.2", output
+    end
+  end
+
+  test "hosts param intersected with configuration" do
+    Kamal::Cli::Accessory.any_instance.expects(:directories).with("redis")
+    Kamal::Cli::Accessory.any_instance.expects(:upload).with("redis")
+
+    run_command("boot", "redis", "--hosts", "1.1.1.1,1.1.1.3").tap do |output|
+      assert_match /docker login.*on 1.1.1.1/, output
+      refute_match /docker login.*on 1.1.1.3/, output
+      assert_match "docker run --name app-redis --detach --restart unless-stopped --log-opt max-size=\"10m\" --publish 6379:6379 --env-file .kamal/env/accessories/app-redis.env --volume $PWD/app-redis/data:/data --label service=\"app-redis\" redis:latest on 1.1.1.1", output
+      refute_match "docker run --name app-redis --detach --restart unless-stopped --log-opt max-size=\"10m\" --publish 6379:6379 --env-file .kamal/env/accessories/app-redis.env --volume $PWD/app-redis/data:/data --label service=\"app-redis\" redis:latest on 1.1.1.3", output
+    end
+  end
+
   private
     def run_command(*command)
       stdouted { Kamal::Cli::Accessory.start([*command, "-c", "test/fixtures/deploy_with_accessories.yml"]) }

test/cli/app_test.rb

@@ -159,7 +159,7 @@ class CliAppTest < CliTestCase
 
   test "exec" do
     run_command("exec", "ruby -v").tap do |output|
-      assert_match "docker run --rm dhh/app:latest ruby -v", output
+      assert_match "docker run --rm --env-file .kamal/env/roles/app-web.env dhh/app:latest ruby -v", output
     end
   end
 
@@ -172,7 +172,7 @@ class CliAppTest < CliTestCase
 
   test "exec interactive" do
     SSHKit::Backend::Abstract.any_instance.expects(:exec)
-      .with("ssh -t root@1.1.1.1 'docker run -it --rm --env-file .kamal/env/roles/app-web.env dhh/app:latest ruby -v'")
+      .with("ssh -t root@1.1.1.1 -p 22 'docker run -it --rm --env-file .kamal/env/roles/app-web.env dhh/app:latest ruby -v'")
     run_command("exec", "-i", "ruby -v").tap do |output|
       assert_match "Get most recent version available as an image...", output
       assert_match "Launching interactive command with version latest via SSH from new container on 1.1.1.1...", output
@@ -181,7 +181,7 @@ class CliAppTest < CliTestCase
 
   test "exec interactive with reuse" do
     SSHKit::Backend::Abstract.any_instance.expects(:exec)
-      .with("ssh -t root@1.1.1.1 'docker exec -it app-web-999 ruby -v'")
+      .with("ssh -t root@1.1.1.1 -p 22 'docker exec -it app-web-999 ruby -v'")
     run_command("exec", "-i", "--reuse", "ruby -v").tap do |output|
       assert_match "Get current version of running container...", output
       assert_match "Running docker ps --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --latest --format \"{{.Names}}\" | while read line; do echo ${line#app-web-}; done on 1.1.1.1", output
@@ -210,7 +210,7 @@ class CliAppTest < CliTestCase
 
   test "logs with follow" do
     SSHKit::Backend::Abstract.any_instance.stubs(:exec)
-      .with("ssh -t root@1.1.1.1 'docker ps --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --latest | xargs docker logs --timestamps --tail 10 --follow 2>&1'")
+      .with("ssh -t root@1.1.1.1 -p 22 'docker ps --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --latest | xargs docker logs --timestamps --tail 10 --follow 2>&1'")
 
     assert_match "docker ps --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --latest | xargs docker logs --timestamps --tail 10 --follow 2>&1", run_command("logs", "--follow")
   end

test/cli/build_test.rb

@@ -57,7 +57,7 @@ class CliBuildTest < CliTestCase
     run_command("pull").tap do |output|
       assert_match /docker image rm --force dhh\/app:999/, output
       assert_match /docker pull dhh\/app:999/, output
-      assert_match "docker inspect -f '{{ .Config.Labels.service }}' dhh/app:999 | grep -x app || (echo \"Image dhh/app:999 is missing the `service` label\" && exit 1)", output
+      assert_match "docker inspect -f '{{ .Config.Labels.service }}' dhh/app:999 | grep -x app || (echo \"Image dhh/app:999 is missing the 'service' label\" && exit 1)", output
     end
   end
 

test/cli/healthcheck_test.rb

@@ -65,8 +65,18 @@ class CliHealthcheckTest < CliTestCase
     assert_match "container not ready (unhealthy)", exception.message
   end
   
+  test "raises an exception if primary does not have traefik" do
+    SSHKit::Backend::Abstract.any_instance.expects(:execute).never
+    
+    exception = assert_raises do
+      run_command("perform", config_file: "test/fixtures/deploy_workers_only.yml")
+    end
+
+    assert_equal "The primary host is not configured to run Traefik", exception.message
+  end
+
   private
-    def run_command(*command)
-      stdouted { Kamal::Cli::Healthcheck.start([*command, "-c", "test/fixtures/deploy_with_accessories.yml"]) }
+    def run_command(*command, config_file: "test/fixtures/deploy_with_accessories.yml")
+      stdouted { Kamal::Cli::Healthcheck.start([*command, "-c", config_file]) } 
     end
 end

test/cli/main_test.rb

@@ -2,12 +2,47 @@ require_relative "cli_test_case"
 
 class CliMainTest < CliTestCase
   test "setup" do
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:server:bootstrap")
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:env:push")
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:accessory:boot", [ "all" ])
+    invoke_options = { "config_file" => "test/fixtures/deploy_simple.yml", "version" => "999", "skip_hooks" => false }
+    
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:server:bootstrap", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:env:push", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:accessory:boot", [ "all" ], invoke_options)
     Kamal::Cli::Main.any_instance.expects(:deploy)
 
-    run_command("setup")
+    run_command("setup").tap do |output|
+      assert_match /Ensure Docker is installed.../, output
+      assert_match /Push env files.../, output
+    end
+  end
+
+  test "setup with skip_push" do
+    invoke_options = { "config_file" => "test/fixtures/deploy_simple.yml", "version" => "999", "skip_hooks" => false }
+
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:server:bootstrap", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:env:push", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:accessory:boot", [ "all" ], invoke_options)
+    # deploy
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:registry:login", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:build:pull", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:traefik:boot", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:healthcheck:perform", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:app:stale_containers", [], invoke_options.merge(stop: true))
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:app:boot", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:prune:all", [], invoke_options)
+
+    run_command("setup", "--skip_push").tap do |output|
+      assert_match /Ensure Docker is installed.../, output
+      assert_match /Push env files.../, output
+      # deploy
+      assert_match /Acquiring the deploy lock/, output
+      assert_match /Log into image registry/, output
+      assert_match /Pull app image/, output
+      assert_match /Ensure Traefik is running/, output
+      assert_match /Ensure app can pass healthcheck/, output
+      assert_match /Detect stale containers/, output
+      assert_match /Prune old containers and images/, output
+      assert_match /Releasing the deploy lock/, output
+    end
   end
 
   test "deploy" do
@@ -123,10 +158,33 @@ class CliMainTest < CliTestCase
     end
   end
   
-  test "deploy with missing secrets" do
-    assert_raises(KeyError) do
-      run_command("deploy", config_file: "deploy_with_secrets")
+  test "deploy without healthcheck if primary host doesn't have traefik" do
+    invoke_options = { "config_file" => "test/fixtures/deploy_workers_only.yml", "version" => "999", "skip_hooks" => false }
+
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:healthcheck:perform", [], invoke_options).never
+    
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:registry:login", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:build:deliver", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:traefik:boot", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:app:stale_containers", [], invoke_options.merge(stop: true))
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:app:boot", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:prune:all", [], invoke_options)
+
+    run_command("deploy", config_file: "deploy_workers_only")
   end
+
+  test "deploy with missing secrets" do
+    invoke_options = { "config_file" => "test/fixtures/deploy_with_secrets.yml", "version" => "999", "skip_hooks" => false }
+
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:registry:login", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:build:deliver", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:traefik:boot", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:healthcheck:perform", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:app:stale_containers", [], invoke_options.merge(stop: true))
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:app:boot", [], invoke_options)
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:prune:all", [], invoke_options)
+
+    run_command("deploy", config_file: "deploy_with_secrets")
   end
 
   test "redeploy" do
@@ -275,6 +333,16 @@ class CliMainTest < CliTestCase
     end
   end
 
+  test "config with primary web role override" do
+    run_command("config", config_file: "deploy_primary_web_role_override").tap do |output|
+      config = YAML.load(output)
+
+      assert_equal ["web_chicago", "web_tokyo"], config[:roles]
+      assert_equal ["1.1.1.1", "1.1.1.2", "1.1.1.3", "1.1.1.4"], config[:hosts]
+      assert_equal "1.1.1.3", config[:primary_host]
+    end
+  end
+
   test "config with destination" do
     run_command("config", "-d", "world", config_file: "deploy_for_dest").tap do |output|
       config = YAML.load(output)
@@ -288,6 +356,19 @@ class CliMainTest < CliTestCase
     end
   end
 
+  test "config with aliases" do
+    run_command("config", config_file: "deploy_with_aliases").tap do |output|
+      config = YAML.load(output)
+
+      assert_equal ["web", "web_tokyo", "workers", "workers_tokyo"], config[:roles]
+      assert_equal ["1.1.1.1", "1.1.1.2", "1.1.1.3", "1.1.1.4"], config[:hosts]
+      assert_equal "999", config[:version]
+      assert_equal "registry.digitalocean.com/dhh/app", config[:repository]
+      assert_equal "registry.digitalocean.com/dhh/app:999", config[:absolute_image]
+      assert_equal "app-999", config[:service_with_version]
+    end
+  end
+
   test "init" do
     Pathname.any_instance.expects(:exist?).returns(false).times(3)
     Pathname.any_instance.stubs(:mkpath)
@@ -346,6 +427,20 @@ class CliMainTest < CliTestCase
     run_command("envify")
   end
 
+  test "envify with blank line trimming" do
+    file = <<~EOF
+      HELLO=<%= 'world' %>
+      <% if true -%>
+      KEY=value
+      <% end  -%>
+    EOF
+
+    File.expects(:read).with(".env.erb").returns(file.strip)
+    File.expects(:write).with(".env", "HELLO=world\nKEY=value\n", perm: 0600)
+
+    run_command("envify")
+  end
+
   test "envify with destination" do
     File.expects(:read).with(".env.world.erb").returns("HELLO=<%= 'world' %>")
     File.expects(:write).with(".env.world", "HELLO=world", perm: 0600)
@@ -353,6 +448,14 @@ class CliMainTest < CliTestCase
     run_command("envify", "-d", "world", config_file: "deploy_for_dest")
   end
 
+  test "envify with skip_push" do
+    File.expects(:read).with(".env.erb").returns("HELLO=<%= 'world' %>")
+    File.expects(:write).with(".env", "HELLO=world", perm: 0600)
+
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:env:push").never
+    run_command("envify", "--skip-push")
+  end
+
   test "remove with confirmation" do
     run_command("remove", "-y", config_file: "deploy_with_accessories").tap do |output|
       assert_match /docker container stop traefik/, output

test/cli/prune_test.rb

@@ -20,6 +20,15 @@ class CliPruneTest < CliTestCase
       assert_match /docker ps -q -a --filter label=service=app --filter status=created --filter status=exited --filter status=dead | tail -n +6 | while read container_id; do docker rm $container_id; done on 1.1.1.\d/, output
       assert_match /docker container prune --force --filter label=service=healthcheck-app on 1.1.1.\d/, output
      end
+
+    run_command("containers", "--retain", "10").tap do |output|
+      assert_match /docker ps -q -a --filter label=service=app --filter status=created --filter status=exited --filter status=dead | tail -n +11 | while read container_id; do docker rm $container_id; done on 1.1.1.\d/, output
+      assert_match /docker container prune --force --filter label=service=healthcheck-app on 1.1.1.\d/, output
+    end
+
+    assert_raises(RuntimeError, "retain must be at least 1") do
+      run_command("containers", "--retain", "0")
+    end
   end
 
   private

test/cli/server_test.rb

@@ -10,7 +10,7 @@ class CliServerTest < CliTestCase
 
   test "bootstrap install as non-root user" do
     SSHKit::Backend::Abstract.any_instance.expects(:execute).with(:docker, "-v", raise_on_non_zero_exit: false).returns(false).at_least_once
-    SSHKit::Backend::Abstract.any_instance.expects(:execute).with('[ "${EUID:-$(id -u)}" -eq 0 ]', raise_on_non_zero_exit: false).returns(false).at_least_once
+    SSHKit::Backend::Abstract.any_instance.expects(:execute).with('[ "${EUID:-$(id -u)}" -eq 0 ] || command -v sudo >/dev/null || command -v su >/dev/null', raise_on_non_zero_exit: false).returns(false).at_least_once
     SSHKit::Backend::Abstract.any_instance.expects(:execute).with(:mkdir, "-p", ".kamal").returns("").at_least_once
 
     assert_raise RuntimeError, "Docker is not installed on 1.1.1.1, 1.1.1.3, 1.1.1.4, 1.1.1.2 and can't be automatically installed without having root access and the `curl` command available. Install Docker manually: https://docs.docker.com/engine/install/" do
@@ -20,13 +20,16 @@ class CliServerTest < CliTestCase
 
   test "bootstrap install as root user" do
     SSHKit::Backend::Abstract.any_instance.expects(:execute).with(:docker, "-v", raise_on_non_zero_exit: false).returns(false).at_least_once
-    SSHKit::Backend::Abstract.any_instance.expects(:execute).with('[ "${EUID:-$(id -u)}" -eq 0 ]', raise_on_non_zero_exit: false).returns(true).at_least_once
-    SSHKit::Backend::Abstract.any_instance.expects(:execute).with(:curl, "-fsSL", "https://get.docker.com", "|", :sh).at_least_once
+    SSHKit::Backend::Abstract.any_instance.expects(:execute).with('[ "${EUID:-$(id -u)}" -eq 0 ] || command -v sudo >/dev/null || command -v su >/dev/null', raise_on_non_zero_exit: false).returns(true).at_least_once
+    SSHKit::Backend::Abstract.any_instance.expects(:execute).with(:sh, "-c", "'curl -fsSL https://get.docker.com || wget -O - https://get.docker.com || echo \"exit 1\"'", "|", :sh).at_least_once
     SSHKit::Backend::Abstract.any_instance.expects(:execute).with(:mkdir, "-p", ".kamal").returns("").at_least_once
+    Kamal::Commands::Hook.any_instance.stubs(:hook_exists?).returns(true)
+    SSHKit::Backend::Abstract.any_instance.expects(:execute).with(".kamal/hooks/docker-setup", anything).at_least_once
 
     run_command("bootstrap").tap do |output|
       ("1.1.1.1".."1.1.1.4").map do |host|
         assert_match "Missing Docker on #{host}. Installing…", output
+        assert_match "Running the docker-setup hook", output
       end
     end
   end

test/cli/traefik_test.rb

@@ -4,7 +4,7 @@ class CliTraefikTest < CliTestCase
   test "boot" do
     run_command("boot").tap do |output|
       assert_match "docker login", output
-      assert_match "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" #{Kamal::Commands::Traefik::DEFAULT_IMAGE} --providers.docker --log.level=\"DEBUG\"", output
+      assert_match "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" #{Kamal::Commands::Traefik::DEFAULT_IMAGE} --providers.docker --log.level=\"DEBUG\"", output
     end
   end
 
@@ -14,7 +14,7 @@ class CliTraefikTest < CliTestCase
     run_command("reboot").tap do |output|
       assert_match "docker container stop traefik", output
       assert_match "docker container prune --force --filter label=org.opencontainers.image.title=Traefik", output
-      assert_match "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" #{Kamal::Commands::Traefik::DEFAULT_IMAGE} --providers.docker --log.level=\"DEBUG\"", output
+      assert_match "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" #{Kamal::Commands::Traefik::DEFAULT_IMAGE} --providers.docker --log.level=\"DEBUG\"", output
     end
   end
 
@@ -64,7 +64,7 @@ class CliTraefikTest < CliTestCase
 
   test "logs with follow" do
     SSHKit::Backend::Abstract.any_instance.stubs(:exec)
-      .with("ssh -t root@1.1.1.1 'docker logs traefik --timestamps --tail 10 --follow 2>&1'")
+      .with("ssh -t root@1.1.1.1 -p 22 'docker logs traefik --timestamps --tail 10 --follow 2>&1'")
 
     assert_match "docker logs traefik --timestamps --tail 10 --follow", run_command("logs", "--follow")
   end

test/commander_test.rb

@@ -14,6 +14,20 @@ class CommanderTest < ActiveSupport::TestCase
 
     @kamal.specific_hosts = [ "1.1.1.1", "1.1.1.2" ]
     assert_equal [ "1.1.1.1", "1.1.1.2" ], @kamal.hosts
+
+    @kamal.specific_hosts = [ "1.1.1.1*" ]
+    assert_equal [ "1.1.1.1" ], @kamal.hosts
+
+    @kamal.specific_hosts = [ "1.1.1.*", "*.1.2.*" ]
+    assert_equal [ "1.1.1.1", "1.1.1.2", "1.1.1.3", "1.1.1.4" ], @kamal.hosts
+
+    @kamal.specific_hosts = [ "*" ]
+    assert_equal [ "1.1.1.1", "1.1.1.2", "1.1.1.3", "1.1.1.4" ], @kamal.hosts
+
+    exception = assert_raises(ArgumentError) do
+      @kamal.specific_hosts = [ "*miss" ]
+    end
+    assert_match /hosts match for \*miss/, exception.message
   end
 
   test "filtering hosts by filtering roles" do
@@ -21,6 +35,11 @@ class CommanderTest < ActiveSupport::TestCase
 
     @kamal.specific_roles = [ "web" ]
     assert_equal [ "1.1.1.1", "1.1.1.2" ], @kamal.hosts
+
+    exception = assert_raises(ArgumentError) do
+      @kamal.specific_roles = [ "*miss" ]
+    end
+    assert_match /roles match for \*miss/, exception.message
   end
 
   test "filtering roles" do
@@ -28,6 +47,20 @@ class CommanderTest < ActiveSupport::TestCase
 
     @kamal.specific_roles = [ "workers" ]
     assert_equal [ "workers" ], @kamal.roles.map(&:name)
+
+    @kamal.specific_roles = [ "w*" ]
+    assert_equal [ "web", "workers" ], @kamal.roles.map(&:name)
+
+    @kamal.specific_roles = [ "we*", "*orkers" ]
+    assert_equal [ "web", "workers" ], @kamal.roles.map(&:name)
+
+    @kamal.specific_roles = [ "*" ]
+    assert_equal [ "web", "workers" ], @kamal.roles.map(&:name)
+
+    exception = assert_raises(ArgumentError) do
+      @kamal.specific_roles = [ "*miss" ]
+    end
+    assert_match /roles match for \*miss/, exception.message
   end
 
   test "filtering roles by filtering hosts" do
@@ -50,14 +83,14 @@ class CommanderTest < ActiveSupport::TestCase
   end
 
   test "primary_role" do
-    assert_equal "web", @kamal.primary_role
+    assert_equal "web", @kamal.primary_role.name
     @kamal.specific_roles = "workers"
-    assert_equal "workers", @kamal.primary_role
+    assert_equal "workers", @kamal.primary_role.name
   end
 
   test "roles_on" do
-    assert_equal [ "web" ], @kamal.roles_on("1.1.1.1")
-    assert_equal [ "workers" ], @kamal.roles_on("1.1.1.3")
+    assert_equal [ "web" ], @kamal.roles_on("1.1.1.1").map(&:name)
+    assert_equal [ "workers" ], @kamal.roles_on("1.1.1.3").map(&:name)
   end
 
   test "default group strategy" do
@@ -76,6 +109,21 @@ class CommanderTest < ActiveSupport::TestCase
     assert_equal({ in: :groups, limit: 1, wait: 2 }, @kamal.boot_strategy)
   end
 
+  test "percentage-based group strategy limit is at least 1" do
+    configure_with(:deploy_with_low_percentage_boot_strategy)
+
+    assert_equal({ in: :groups, limit: 1, wait: 2 }, @kamal.boot_strategy)
+  end
+
+  test "try to match the primary role from a list of specific roles" do
+    configure_with(:deploy_primary_web_role_override)
+
+    @kamal.specific_roles = [ "web_*" ]
+    assert_equal [ "web_chicago", "web_tokyo" ], @kamal.roles.map(&:name)
+    assert_equal "web_tokyo", @kamal.primary_role.name
+    assert_equal "1.1.1.3", @kamal.primary_host
+  end
+
   private
     def configure_with(variant)
       @kamal = Kamal::Commander.new.tap do |kamal|

test/commands/accessory_test.rb

@@ -34,6 +34,7 @@ class CommandsAccessoryTest < ActiveSupport::TestCase
           ]
         },
         "busybox" => {
+          "service" => "custom-busybox",
           "image" => "busybox:latest",
           "host" => "1.1.1.7"
         }
@@ -57,7 +58,7 @@ class CommandsAccessoryTest < ActiveSupport::TestCase
       new_command(:redis).run.join(" ")
 
     assert_equal \
-      "docker run --name app-busybox --detach --restart unless-stopped --log-opt max-size=\"10m\" --env-file .kamal/env/accessories/app-busybox.env --label service=\"app-busybox\" busybox:latest",
+      "docker run --name custom-busybox --detach --restart unless-stopped --log-opt max-size=\"10m\" --env-file .kamal/env/accessories/custom-busybox.env --label service=\"custom-busybox\" busybox:latest",
       new_command(:busybox).run.join(" ")
   end
 
@@ -65,7 +66,7 @@ class CommandsAccessoryTest < ActiveSupport::TestCase
     @config[:logging] = { "driver" => "local", "options" => { "max-size" => "100m", "max-file" => "3" } }
 
     assert_equal \
-      "docker run --name app-busybox --detach --restart unless-stopped --log-driver \"local\" --log-opt max-size=\"100m\" --log-opt max-file=\"3\" --env-file .kamal/env/accessories/app-busybox.env --label service=\"app-busybox\" busybox:latest",
+      "docker run --name custom-busybox --detach --restart unless-stopped --log-driver \"local\" --log-opt max-size=\"100m\" --log-opt max-file=\"3\" --env-file .kamal/env/accessories/custom-busybox.env --label service=\"custom-busybox\" busybox:latest",
       new_command(:busybox).run.join(" ")
   end
 
@@ -128,7 +129,7 @@ class CommandsAccessoryTest < ActiveSupport::TestCase
 
   test "follow logs" do
     assert_equal \
-      "ssh -t root@1.1.1.5 'docker logs app-mysql --timestamps --tail 10 --follow 2>&1'",
+      "ssh -t root@1.1.1.5 -p 22 'docker logs app-mysql --timestamps --tail 10 --follow 2>&1'",
       new_command(:mysql).follow_logs
   end
 

test/commands/app_test.rb

@@ -14,13 +14,13 @@ class CommandsAppTest < ActiveSupport::TestCase
 
   test "run" do
     assert_equal \
-      "docker run --detach --restart unless-stopped --name app-web-999 -e KAMAL_CONTAINER_NAME=\"app-web-999\" --env-file .kamal/env/roles/app-web.env --health-cmd \"(curl -f http://localhost:3000/up || exit 1) && (stat /tmp/kamal-cord/cord > /dev/null || exit 1)\" --health-interval \"1s\" --volume $(pwd)/.kamal/cords/app-web-12345678901234567890123456789012:/tmp/kamal-cord --log-opt max-size=\"10m\" --label service=\"app\" --label role=\"web\" --label traefik.http.services.app-web.loadbalancer.server.scheme=\"http\" --label traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.middlewares.app-web-retry.retry.attempts=\"5\" --label traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\" --label traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\" dhh/app:999",
+      "docker run --detach --restart unless-stopped --name app-web-999 -e KAMAL_CONTAINER_NAME=\"app-web-999\" -e KAMAL_VERSION=\"999\" --env-file .kamal/env/roles/app-web.env --health-cmd \"(curl -f http://localhost:3000/up || exit 1) && (stat /tmp/kamal-cord/cord > /dev/null || exit 1)\" --health-interval \"1s\" --volume $(pwd)/.kamal/cords/app-web-12345678901234567890123456789012:/tmp/kamal-cord --log-opt max-size=\"10m\" --label service=\"app\" --label role=\"web\" --label traefik.http.services.app-web.loadbalancer.server.scheme=\"http\" --label traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.app-web.priority=\"2\" --label traefik.http.middlewares.app-web-retry.retry.attempts=\"5\" --label traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\" --label traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\" dhh/app:999",
       new_command.run.join(" ")
   end
 
   test "run with hostname" do
     assert_equal \
-      "docker run --detach --restart unless-stopped --name app-web-999 --hostname myhost -e KAMAL_CONTAINER_NAME=\"app-web-999\" --env-file .kamal/env/roles/app-web.env --health-cmd \"(curl -f http://localhost:3000/up || exit 1) && (stat /tmp/kamal-cord/cord > /dev/null || exit 1)\" --health-interval \"1s\" --volume $(pwd)/.kamal/cords/app-web-12345678901234567890123456789012:/tmp/kamal-cord --log-opt max-size=\"10m\" --label service=\"app\" --label role=\"web\" --label traefik.http.services.app-web.loadbalancer.server.scheme=\"http\" --label traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.middlewares.app-web-retry.retry.attempts=\"5\" --label traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\" --label traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\" dhh/app:999",
+      "docker run --detach --restart unless-stopped --name app-web-999 --hostname myhost -e KAMAL_CONTAINER_NAME=\"app-web-999\" -e KAMAL_VERSION=\"999\" --env-file .kamal/env/roles/app-web.env --health-cmd \"(curl -f http://localhost:3000/up || exit 1) && (stat /tmp/kamal-cord/cord > /dev/null || exit 1)\" --health-interval \"1s\" --volume $(pwd)/.kamal/cords/app-web-12345678901234567890123456789012:/tmp/kamal-cord --log-opt max-size=\"10m\" --label service=\"app\" --label role=\"web\" --label traefik.http.services.app-web.loadbalancer.server.scheme=\"http\" --label traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.app-web.priority=\"2\" --label traefik.http.middlewares.app-web-retry.retry.attempts=\"5\" --label traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\" --label traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\" dhh/app:999",
       new_command.run(hostname: "myhost").join(" ")
   end
 
@@ -28,7 +28,7 @@ class CommandsAppTest < ActiveSupport::TestCase
     @config[:volumes] = ["/local/path:/container/path" ]
 
     assert_equal \
-      "docker run --detach --restart unless-stopped --name app-web-999 -e KAMAL_CONTAINER_NAME=\"app-web-999\" --env-file .kamal/env/roles/app-web.env --health-cmd \"(curl -f http://localhost:3000/up || exit 1) && (stat /tmp/kamal-cord/cord > /dev/null || exit 1)\" --health-interval \"1s\" --volume $(pwd)/.kamal/cords/app-web-12345678901234567890123456789012:/tmp/kamal-cord --log-opt max-size=\"10m\" --volume /local/path:/container/path --label service=\"app\" --label role=\"web\" --label traefik.http.services.app-web.loadbalancer.server.scheme=\"http\" --label traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.middlewares.app-web-retry.retry.attempts=\"5\" --label traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\" --label traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\" dhh/app:999",
+      "docker run --detach --restart unless-stopped --name app-web-999 -e KAMAL_CONTAINER_NAME=\"app-web-999\" -e KAMAL_VERSION=\"999\" --env-file .kamal/env/roles/app-web.env --health-cmd \"(curl -f http://localhost:3000/up || exit 1) && (stat /tmp/kamal-cord/cord > /dev/null || exit 1)\" --health-interval \"1s\" --volume $(pwd)/.kamal/cords/app-web-12345678901234567890123456789012:/tmp/kamal-cord --log-opt max-size=\"10m\" --volume /local/path:/container/path --label service=\"app\" --label role=\"web\" --label traefik.http.services.app-web.loadbalancer.server.scheme=\"http\" --label traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.app-web.priority=\"2\" --label traefik.http.middlewares.app-web-retry.retry.attempts=\"5\" --label traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\" --label traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\" dhh/app:999",
       new_command.run.join(" ")
   end
 
@@ -36,7 +36,7 @@ class CommandsAppTest < ActiveSupport::TestCase
     @config[:healthcheck] = { "path" => "/healthz" }
 
     assert_equal \
-      "docker run --detach --restart unless-stopped --name app-web-999 -e KAMAL_CONTAINER_NAME=\"app-web-999\" --env-file .kamal/env/roles/app-web.env --health-cmd \"(curl -f http://localhost:3000/healthz || exit 1) && (stat /tmp/kamal-cord/cord > /dev/null || exit 1)\" --health-interval \"1s\" --volume $(pwd)/.kamal/cords/app-web-12345678901234567890123456789012:/tmp/kamal-cord --log-opt max-size=\"10m\" --label service=\"app\" --label role=\"web\" --label traefik.http.services.app-web.loadbalancer.server.scheme=\"http\" --label traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.middlewares.app-web-retry.retry.attempts=\"5\" --label traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\" --label traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\" dhh/app:999",
+      "docker run --detach --restart unless-stopped --name app-web-999 -e KAMAL_CONTAINER_NAME=\"app-web-999\" -e KAMAL_VERSION=\"999\" --env-file .kamal/env/roles/app-web.env --health-cmd \"(curl -f http://localhost:3000/healthz || exit 1) && (stat /tmp/kamal-cord/cord > /dev/null || exit 1)\" --health-interval \"1s\" --volume $(pwd)/.kamal/cords/app-web-12345678901234567890123456789012:/tmp/kamal-cord --log-opt max-size=\"10m\" --label service=\"app\" --label role=\"web\" --label traefik.http.services.app-web.loadbalancer.server.scheme=\"http\" --label traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.app-web.priority=\"2\" --label traefik.http.middlewares.app-web-retry.retry.attempts=\"5\" --label traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\" --label traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\" dhh/app:999",
       new_command.run.join(" ")
   end
 
@@ -44,7 +44,7 @@ class CommandsAppTest < ActiveSupport::TestCase
     @config[:healthcheck] = { "cmd" => "/bin/up" }
 
     assert_equal \
-      "docker run --detach --restart unless-stopped --name app-web-999 -e KAMAL_CONTAINER_NAME=\"app-web-999\" --env-file .kamal/env/roles/app-web.env --health-cmd \"(/bin/up) && (stat /tmp/kamal-cord/cord > /dev/null || exit 1)\" --health-interval \"1s\" --volume $(pwd)/.kamal/cords/app-web-12345678901234567890123456789012:/tmp/kamal-cord --log-opt max-size=\"10m\" --label service=\"app\" --label role=\"web\" --label traefik.http.services.app-web.loadbalancer.server.scheme=\"http\" --label traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.middlewares.app-web-retry.retry.attempts=\"5\" --label traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\" --label traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\" dhh/app:999",
+      "docker run --detach --restart unless-stopped --name app-web-999 -e KAMAL_CONTAINER_NAME=\"app-web-999\" -e KAMAL_VERSION=\"999\" --env-file .kamal/env/roles/app-web.env --health-cmd \"(/bin/up) && (stat /tmp/kamal-cord/cord > /dev/null || exit 1)\" --health-interval \"1s\" --volume $(pwd)/.kamal/cords/app-web-12345678901234567890123456789012:/tmp/kamal-cord --log-opt max-size=\"10m\" --label service=\"app\" --label role=\"web\" --label traefik.http.services.app-web.loadbalancer.server.scheme=\"http\" --label traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.app-web.priority=\"2\" --label traefik.http.middlewares.app-web-retry.retry.attempts=\"5\" --label traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\" --label traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\" dhh/app:999",
       new_command.run.join(" ")
   end
 
@@ -52,14 +52,14 @@ class CommandsAppTest < ActiveSupport::TestCase
     @config[:servers] = { "web" => { "hosts" => [ "1.1.1.1" ], "healthcheck" => { "cmd" => "/bin/healthy" } } }
 
     assert_equal \
-      "docker run --detach --restart unless-stopped --name app-web-999 -e KAMAL_CONTAINER_NAME=\"app-web-999\" --env-file .kamal/env/roles/app-web.env --health-cmd \"(/bin/healthy) && (stat /tmp/kamal-cord/cord > /dev/null || exit 1)\" --health-interval \"1s\" --volume $(pwd)/.kamal/cords/app-web-12345678901234567890123456789012:/tmp/kamal-cord --log-opt max-size=\"10m\" --label service=\"app\" --label role=\"web\" --label traefik.http.services.app-web.loadbalancer.server.scheme=\"http\" --label traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.middlewares.app-web-retry.retry.attempts=\"5\" --label traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\" --label traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\" dhh/app:999",
+      "docker run --detach --restart unless-stopped --name app-web-999 -e KAMAL_CONTAINER_NAME=\"app-web-999\" -e KAMAL_VERSION=\"999\" --env-file .kamal/env/roles/app-web.env --health-cmd \"(/bin/healthy) && (stat /tmp/kamal-cord/cord > /dev/null || exit 1)\" --health-interval \"1s\" --volume $(pwd)/.kamal/cords/app-web-12345678901234567890123456789012:/tmp/kamal-cord --log-opt max-size=\"10m\" --label service=\"app\" --label role=\"web\" --label traefik.http.services.app-web.loadbalancer.server.scheme=\"http\" --label traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.app-web.priority=\"2\" --label traefik.http.middlewares.app-web-retry.retry.attempts=\"5\" --label traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\" --label traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\" dhh/app:999",
       new_command.run.join(" ")
   end
 
   test "run with custom options" do
     @config[:servers] = { "web" => [ "1.1.1.1" ], "jobs" => { "hosts" => [ "1.1.1.2" ], "cmd" => "bin/jobs", "options" => { "mount" => "somewhere", "cap-add" => true } } }
     assert_equal \
-      "docker run --detach --restart unless-stopped --name app-jobs-999 -e KAMAL_CONTAINER_NAME=\"app-jobs-999\" --env-file .kamal/env/roles/app-jobs.env --log-opt max-size=\"10m\" --label service=\"app\" --label role=\"jobs\" --mount \"somewhere\" --cap-add dhh/app:999 bin/jobs",
+      "docker run --detach --restart unless-stopped --name app-jobs-999 -e KAMAL_CONTAINER_NAME=\"app-jobs-999\" -e KAMAL_VERSION=\"999\" --env-file .kamal/env/roles/app-jobs.env --log-opt max-size=\"10m\" --label service=\"app\" --label role=\"jobs\" --mount \"somewhere\" --cap-add dhh/app:999 bin/jobs",
       new_command(role: "jobs").run.join(" ")
   end
 
@@ -67,7 +67,16 @@ class CommandsAppTest < ActiveSupport::TestCase
     @config[:logging] = { "driver" => "local", "options" => { "max-size" => "100m", "max-file" => "3" } }
 
     assert_equal \
-      "docker run --detach --restart unless-stopped --name app-web-999 -e KAMAL_CONTAINER_NAME=\"app-web-999\" --env-file .kamal/env/roles/app-web.env --health-cmd \"(curl -f http://localhost:3000/up || exit 1) && (stat /tmp/kamal-cord/cord > /dev/null || exit 1)\" --health-interval \"1s\" --volume $(pwd)/.kamal/cords/app-web-12345678901234567890123456789012:/tmp/kamal-cord --log-driver \"local\" --log-opt max-size=\"100m\" --log-opt max-file=\"3\" --label service=\"app\" --label role=\"web\" --label traefik.http.services.app-web.loadbalancer.server.scheme=\"http\" --label traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.middlewares.app-web-retry.retry.attempts=\"5\" --label traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\" --label traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\" dhh/app:999",
+      "docker run --detach --restart unless-stopped --name app-web-999 -e KAMAL_CONTAINER_NAME=\"app-web-999\" -e KAMAL_VERSION=\"999\" --env-file .kamal/env/roles/app-web.env --health-cmd \"(curl -f http://localhost:3000/up || exit 1) && (stat /tmp/kamal-cord/cord > /dev/null || exit 1)\" --health-interval \"1s\" --volume $(pwd)/.kamal/cords/app-web-12345678901234567890123456789012:/tmp/kamal-cord --log-driver \"local\" --log-opt max-size=\"100m\" --log-opt max-file=\"3\" --label service=\"app\" --label role=\"web\" --label traefik.http.services.app-web.loadbalancer.server.scheme=\"http\" --label traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.app-web.priority=\"2\" --label traefik.http.middlewares.app-web-retry.retry.attempts=\"5\" --label traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\" --label traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\" dhh/app:999",
+      new_command.run.join(" ")
+  end
+
+  test "run with role logging config" do
+    @config[:logging] = { "driver" => "local", "options" => { "max-size" => "10m", "max-file" => "3" } }
+    @config[:servers] = { "web" => { "hosts" => [ "1.1.1.1" ], "logging" => { "driver" => "local", "options" => { "max-size" => "100m" } } } }
+
+    assert_equal \
+      "docker run --detach --restart unless-stopped --name app-web-999 -e KAMAL_CONTAINER_NAME=\"app-web-999\" -e KAMAL_VERSION=\"999\" --env-file .kamal/env/roles/app-web.env --health-cmd \"(curl -f http://localhost:3000/up || exit 1) && (stat /tmp/kamal-cord/cord > /dev/null || exit 1)\" --health-interval \"1s\" --volume $(pwd)/.kamal/cords/app-web-12345678901234567890123456789012:/tmp/kamal-cord --log-driver \"local\" --log-opt max-size=\"100m\" --log-opt max-file=\"3\" --label service=\"app\" --label role=\"web\" --label traefik.http.services.app-web.loadbalancer.server.scheme=\"http\" --label traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.app-web.priority=\"2\" --label traefik.http.middlewares.app-web-retry.retry.attempts=\"5\" --label traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\" --label traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\" dhh/app:999",
       new_command.run.join(" ")
   end
 
@@ -145,12 +154,20 @@ class CommandsAppTest < ActiveSupport::TestCase
 
   test "follow logs" do
     assert_match \
-      "docker ps --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --latest | xargs docker logs --timestamps --tail 10 --follow 2>&1",
+      "docker ps --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --latest | xargs docker logs --timestamps --follow 2>&1",
       new_command.follow_logs(host: "app-1")
 
     assert_match \
-      "docker ps --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --latest | xargs docker logs --timestamps --tail 10 --follow 2>&1 | grep \"Completed\"",
+      "docker ps --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --latest | xargs docker logs --timestamps --follow 2>&1 | grep \"Completed\"",
       new_command.follow_logs(host: "app-1", grep: "Completed")
+
+    assert_match \
+      "docker ps --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --latest | xargs docker logs --timestamps --tail 123 --follow 2>&1",
+      new_command.follow_logs(host: "app-1", lines: 123)
+
+    assert_match \
+      "docker ps --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --latest | xargs docker logs --timestamps --tail 123 --follow 2>&1 | grep \"Completed\"",
+      new_command.follow_logs(host: "app-1", lines: 123, grep: "Completed")
   end
 
 
@@ -190,32 +207,37 @@ class CommandsAppTest < ActiveSupport::TestCase
   end
 
   test "run over ssh" do
-    assert_equal "ssh -t root@1.1.1.1 'ls'", new_command.run_over_ssh("ls", host: "1.1.1.1")
+    assert_equal "ssh -t root@1.1.1.1 -p 22 'ls'", new_command.run_over_ssh("ls", host: "1.1.1.1")
   end
 
   test "run over ssh with custom user" do
     @config[:ssh] = { "user" => "app" }
-    assert_equal "ssh -t app@1.1.1.1 'ls'", new_command.run_over_ssh("ls", host: "1.1.1.1")
+    assert_equal "ssh -t app@1.1.1.1 -p 22 'ls'", new_command.run_over_ssh("ls", host: "1.1.1.1")
+  end
+
+  test "run over ssh with custom port" do
+    @config[:ssh] = { "port" => "2222" }
+    assert_equal "ssh -t root@1.1.1.1 -p 2222 'ls'", new_command.run_over_ssh("ls", host: "1.1.1.1")
   end
 
   test "run over ssh with proxy" do
     @config[:ssh] = { "proxy" => "2.2.2.2" }
-    assert_equal "ssh -J root@2.2.2.2 -t root@1.1.1.1 'ls'", new_command.run_over_ssh("ls", host: "1.1.1.1")
+    assert_equal "ssh -J root@2.2.2.2 -t root@1.1.1.1 -p 22 'ls'", new_command.run_over_ssh("ls", host: "1.1.1.1")
   end
 
   test "run over ssh with proxy user" do
     @config[:ssh] = { "proxy" => "app@2.2.2.2" }
-    assert_equal "ssh -J app@2.2.2.2 -t root@1.1.1.1 'ls'", new_command.run_over_ssh("ls", host: "1.1.1.1")
+    assert_equal "ssh -J app@2.2.2.2 -t root@1.1.1.1 -p 22 'ls'", new_command.run_over_ssh("ls", host: "1.1.1.1")
   end
 
   test "run over ssh with custom user with proxy" do
     @config[:ssh] = { "user" => "app", "proxy" => "2.2.2.2" }
-    assert_equal "ssh -J root@2.2.2.2 -t app@1.1.1.1 'ls'", new_command.run_over_ssh("ls", host: "1.1.1.1")
+    assert_equal "ssh -J root@2.2.2.2 -t app@1.1.1.1 -p 22 'ls'", new_command.run_over_ssh("ls", host: "1.1.1.1")
   end
 
   test "run over ssh with proxy_command" do
     @config[:ssh] = { "proxy_command" => "ssh -W %h:%p user@proxy-server" }
-    assert_equal "ssh -o ProxyCommand='ssh -W %h:%p user@proxy-server' -t root@1.1.1.1 'ls'", new_command.run_over_ssh("ls", host: "1.1.1.1")
+    assert_equal "ssh -o ProxyCommand='ssh -W %h:%p user@proxy-server' -t root@1.1.1.1 -p 22 'ls'", new_command.run_over_ssh("ls", host: "1.1.1.1")
   end
 
   test "current_running_container_id" do
@@ -378,6 +400,7 @@ class CommandsAppTest < ActiveSupport::TestCase
 
   private
     def new_command(role: "web", **additional_config)
-      Kamal::Commands::App.new(Kamal::Configuration.new(@config.merge(additional_config), destination: @destination, version: "999"), role: role)
+      config = Kamal::Configuration.new(@config.merge(additional_config), destination: @destination, version: "999")
+      Kamal::Commands::App.new(config, role: config.role(role))
     end
 end

test/commands/builder_test.rb

@@ -37,6 +37,14 @@ class CommandsBuilderTest < ActiveSupport::TestCase
       builder.push.join(" ")
   end
 
+  test "target multiarch local when arch is set" do
+    builder = new_builder_command(builder: { "local" => { "arch" => "amd64" } })
+    assert_equal "multiarch", builder.name
+    assert_equal \
+      "docker buildx build --push --platform linux/amd64 --builder kamal-app-multiarch -t dhh/app:123 -t dhh/app:latest --label service=\"app\" --file Dockerfile .",
+      builder.push.join(" ")
+  end
+
   test "target native remote when only remote is set" do
     builder = new_builder_command(builder: { "remote" => { "arch" => "amd64" }, "cache" => { "type" => "gha" } })
     assert_equal "native/remote", builder.name
@@ -103,8 +111,16 @@ class CommandsBuilderTest < ActiveSupport::TestCase
       builder.push.join(" ")
   end
 
+  test "build with ssh agent socket" do
+    builder = new_builder_command(builder: { "ssh" => 'default=$SSH_AUTH_SOCK' })
+
+    assert_equal \
+      "-t dhh/app:123 -t dhh/app:latest --label service=\"app\" --file Dockerfile --ssh default=$SSH_AUTH_SOCK",
+      builder.target.build_options.join(" ")
+  end
+
   test "validate image" do
-    assert_equal "docker inspect -f '{{ .Config.Labels.service }}' dhh/app:123 | grep -x app || (echo \"Image dhh/app:123 is missing the `service` label\" && exit 1)", new_builder_command.validate_image.join(" ")
+    assert_equal "docker inspect -f '{{ .Config.Labels.service }}' dhh/app:123 | grep -x app || (echo \"Image dhh/app:123 is missing the 'service' label\" && exit 1)", new_builder_command.validate_image.join(" ")
   end
 
   private

test/commands/docker_test.rb

@@ -9,7 +9,7 @@ class CommandsDockerTest < ActiveSupport::TestCase
   end
 
   test "install" do
-    assert_equal "curl -fsSL https://get.docker.com | sh", @docker.install.join(" ")
+    assert_equal "sh -c 'curl -fsSL https://get.docker.com || wget -O - https://get.docker.com || echo \"exit 1\"' | sh", @docker.install.join(" ")
   end
 
   test "installed?" do
@@ -21,6 +21,6 @@ class CommandsDockerTest < ActiveSupport::TestCase
   end
 
   test "superuser?" do
-    assert_equal '[ "${EUID:-$(id -u)}" -eq 0 ]', @docker.superuser?.join(" ")
+    assert_equal '[ "${EUID:-$(id -u)}" -eq 0 ] || command -v sudo >/dev/null || command -v su >/dev/null', @docker.superuser?.join(" ")
   end
 end

test/commands/prune_test.rb

@@ -23,7 +23,11 @@ class CommandsPruneTest < ActiveSupport::TestCase
   test "app containers" do
     assert_equal \
       "docker ps -q -a --filter label=service=app --filter status=created --filter status=exited --filter status=dead | tail -n +6 | while read container_id; do docker rm $container_id; done",
-      new_command.app_containers.join(" ")
+      new_command.app_containers(retain: 5).join(" ")
+
+    assert_equal \
+      "docker ps -q -a --filter label=service=app --filter status=created --filter status=exited --filter status=dead | tail -n +4 | while read container_id; do docker rm $container_id; done",
+      new_command.app_containers(retain: 3).join(" ")
   end
 
   test "healthcheck containers" do

test/commands/registry_test.rb

@@ -15,7 +15,7 @@ class CommandsRegistryTest < ActiveSupport::TestCase
 
   test "registry login" do
     assert_equal \
-      "docker login hub.docker.com -u dhh -p secret",
+      "docker login hub.docker.com -u \"dhh\" -p \"secret\"",
       @registry.login.join(" ")
   end
 
@@ -24,7 +24,18 @@ class CommandsRegistryTest < ActiveSupport::TestCase
     @config[:registry]["password"] = [ "KAMAL_REGISTRY_PASSWORD" ]
 
     assert_equal \
-      "docker login hub.docker.com -u dhh -p more-secret",
+      "docker login hub.docker.com -u \"dhh\" -p \"more-secret\"",
+      @registry.login.join(" ")
+  ensure
+    ENV.delete("KAMAL_REGISTRY_PASSWORD")
+  end
+
+  test "registry login escape password" do
+    ENV["KAMAL_REGISTRY_PASSWORD"] = "more-secret'\""
+    @config[:registry]["password"] = [ "KAMAL_REGISTRY_PASSWORD" ]
+
+    assert_equal \
+      "docker login hub.docker.com -u \"dhh\" -p \"more-secret'\\\"\"",
       @registry.login.join(" ")
   ensure
     ENV.delete("KAMAL_REGISTRY_PASSWORD")
@@ -35,7 +46,7 @@ class CommandsRegistryTest < ActiveSupport::TestCase
     @config[:registry]["username"] = [ "KAMAL_REGISTRY_USERNAME" ]
 
     assert_equal \
-      "docker login hub.docker.com -u also-secret -p secret",
+      "docker login hub.docker.com -u \"also-secret\" -p \"secret\"",
       @registry.login.join(" ")
   ensure
     ENV.delete("KAMAL_REGISTRY_USERNAME")

test/commands/traefik_test.rb

@@ -18,72 +18,72 @@ class CommandsTraefikTest < ActiveSupport::TestCase
 
   test "run" do
     assert_equal \
-      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
+      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
       new_command.run.join(" ")
 
     @config[:traefik]["host_port"] = "8080"
     assert_equal \
-      "docker run --name traefik --detach --restart unless-stopped --publish 8080:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
+      "docker run --name traefik --detach --restart unless-stopped --publish 8080:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
       new_command.run.join(" ")
 
     @config[:traefik]["publish"] = false
     assert_equal \
-      "docker run --name traefik --detach --restart unless-stopped --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
+      "docker run --name traefik --detach --restart unless-stopped --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
       new_command.run.join(" ")
   end
 
   test "run with ports configured" do
     assert_equal \
-      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
+      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
       new_command.run.join(" ")
 
     @config[:traefik]["options"] = {"publish" => %w[9000:9000 9001:9001]}
     assert_equal \
-      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --publish \"9000:9000\" --publish \"9001:9001\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
+      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" --publish \"9000:9000\" --publish \"9001:9001\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
       new_command.run.join(" ")
   end
 
   test "run with volumes configured" do
     assert_equal \
-      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
+      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
       new_command.run.join(" ")
 
     @config[:traefik]["options"] = {"volume" => %w[./letsencrypt/acme.json:/letsencrypt/acme.json] }
     assert_equal \
-      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --volume \"./letsencrypt/acme.json:/letsencrypt/acme.json\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
+      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" --volume \"./letsencrypt/acme.json:/letsencrypt/acme.json\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
       new_command.run.join(" ")
   end
 
   test "run with several options configured" do
     assert_equal \
-      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
+      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
       new_command.run.join(" ")
 
     @config[:traefik]["options"] = {"volume" => %w[./letsencrypt/acme.json:/letsencrypt/acme.json], "publish" => %w[8080:8080], "memory" => "512m"}
     assert_equal \
-      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --volume \"./letsencrypt/acme.json:/letsencrypt/acme.json\" --publish \"8080:8080\" --memory \"512m\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
+      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" --volume \"./letsencrypt/acme.json:/letsencrypt/acme.json\" --publish \"8080:8080\" --memory \"512m\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
       new_command.run.join(" ")
   end
 
   test "run with labels configured" do
     assert_equal \
-      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
+      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
       new_command.run.join(" ")
 
     @config[:traefik]["labels"] = { "traefik.http.routers.dashboard.service" => "api@internal", "traefik.http.routers.dashboard.middlewares" => "auth" }
     assert_equal \
-      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.dashboard.service=\"api@internal\" --label traefik.http.routers.dashboard.middlewares=\"auth\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
+      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" --label traefik.http.routers.dashboard.service=\"api@internal\" --label traefik.http.routers.dashboard.middlewares=\"auth\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
       new_command.run.join(" ")
   end
 
   test "run with env configured" do
     assert_equal \
-      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
+      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
       new_command.run.join(" ")
 
     @config[:traefik]["env"] = { "secret" => %w[EXAMPLE_API_KEY] }
     assert_equal \
-      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
+      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
       new_command.run.join(" ")
   end
 
@@ -91,7 +91,7 @@ class CommandsTraefikTest < ActiveSupport::TestCase
     @config.delete(:traefik)
 
     assert_equal \
-      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" #{Kamal::Commands::Traefik::DEFAULT_IMAGE} --providers.docker --log.level=\"DEBUG\"",
+      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" #{Kamal::Commands::Traefik::DEFAULT_IMAGE} --providers.docker --log.level=\"DEBUG\"",
       new_command.run.join(" ")
   end
 
@@ -99,7 +99,7 @@ class CommandsTraefikTest < ActiveSupport::TestCase
     @config[:logging] = { "driver" => "local", "options" => { "max-size" => "100m", "max-file" => "3" } }
 
     assert_equal \
-      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-driver \"local\" --log-opt max-size=\"100m\" --log-opt max-file=\"3\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
+      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-driver \"local\" --log-opt max-size=\"100m\" --log-opt max-file=\"3\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" #{@image} --providers.docker --log.level=\"DEBUG\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
       new_command.run.join(" ")
   end
 
@@ -107,7 +107,7 @@ class CommandsTraefikTest < ActiveSupport::TestCase
     @config[:traefik]["args"]["log.level"] = "ERROR"
 
     assert_equal \
-    "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" #{@image} --providers.docker --log.level=\"ERROR\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
+    "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env-file .kamal/env/traefik/traefik.env --log-opt max-size=\"10m\" --label traefik.http.routers.catchall.entryPoints=\"http\" --label traefik.http.routers.catchall.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.routers.catchall.service=\"unavailable\" --label traefik.http.routers.catchall.priority=\"1\" --label traefik.http.services.unavailable.loadbalancer.server.port=\"0\" #{@image} --providers.docker --log.level=\"ERROR\" --accesslog.format=\"json\" --api.insecure --metrics.prometheus.buckets=\"0.1,0.3,1.2,5.0\"",
       new_command.run.join(" ")
   end
 
@@ -167,13 +167,13 @@ class CommandsTraefikTest < ActiveSupport::TestCase
 
   test "traefik follow logs" do
     assert_equal \
-      "ssh -t root@1.1.1.1 'docker logs traefik --timestamps --tail 10 --follow 2>&1'",
+      "ssh -t root@1.1.1.1 -p 22 'docker logs traefik --timestamps --tail 10 --follow 2>&1'",
       new_command.follow_logs(host: @config[:servers].first)
   end
 
   test "traefik follow logs with grep hello!" do
     assert_equal \
-      "ssh -t root@1.1.1.1 'docker logs traefik --timestamps --tail 10 --follow 2>&1 | grep \"hello!\"'",
+      "ssh -t root@1.1.1.1 -p 22 'docker logs traefik --timestamps --tail 10 --follow 2>&1 | grep \"hello!\"'",
       new_command.follow_logs(host: @config[:servers].first, grep: 'hello!')
   end
 

test/configuration/accessory_test.rb

@@ -49,6 +49,7 @@ class ConfigurationAccessoryTest < ActiveSupport::TestCase
           }
         },
         "monitoring" => {
+          "service" => "custom-monitoring",
           "image" => "monitoring:latest",
           "roles" => [ "web" ],
           "port" => "4321:4321",
@@ -72,6 +73,7 @@ class ConfigurationAccessoryTest < ActiveSupport::TestCase
   test "service name" do
     assert_equal "app-mysql", @config.accessory(:mysql).service_name
     assert_equal "app-redis", @config.accessory(:redis).service_name
+    assert_equal "custom-monitoring", @config.accessory(:monitoring).service_name
   end
 
   test "port" do
@@ -149,10 +151,16 @@ class ConfigurationAccessoryTest < ActiveSupport::TestCase
     assert_match "%", @config.accessory(:mysql).files.keys[2].read
   end
 
-  test "directories" do
+  test "directory with a relative path" do
+    @deploy[:accessories]["mysql"]["directories"] = [ "data:/var/lib/mysql" ]
     assert_equal({"$PWD/app-mysql/data"=>"/var/lib/mysql"}, @config.accessory(:mysql).directories)
   end
 
+  test "directory with an absolute path" do
+    @deploy[:accessories]["mysql"]["directories"] = [ "/var/data/mysql:/var/lib/mysql" ]
+    assert_equal({"/var/data/mysql"=>"/var/lib/mysql"}, @config.accessory(:mysql).directories)
+  end
+
   test "options" do
     assert_equal ["--cpus", "\"4\"", "--memory", "\"2GB\""], @config.accessory(:redis).option_args
   end

test/configuration/builder_test.rb

@@ -148,4 +148,14 @@ class ConfigurationBuilderTest < ActiveSupport::TestCase
 
     assert_equal "..", @config_with_builder_option.builder.context
   end
+
+  test "ssh" do
+    assert_nil @config.builder.ssh
+  end
+
+  test "setting ssh params" do
+    @deploy_with_builder_option[:builder] = { "ssh" => 'default=$SSH_AUTH_SOCK' }
+
+    assert_equal 'default=$SSH_AUTH_SOCK', @config_with_builder_option.builder.ssh
+  end
 end

test/configuration/role_test.rb

@@ -42,7 +42,7 @@ class ConfigurationRoleTest < ActiveSupport::TestCase
   end
 
   test "special label args for web" do
-    assert_equal [ "--label", "service=\"app\"", "--label", "role=\"web\"", "--label", "traefik.http.services.app-web.loadbalancer.server.scheme=\"http\"", "--label", "traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\"", "--label", "traefik.http.middlewares.app-web-retry.retry.attempts=\"5\"", "--label", "traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\"", "--label", "traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\"" ], @config.role(:web).label_args
+    assert_equal [ "--label", "service=\"app\"", "--label", "role=\"web\"", "--label", "traefik.http.services.app-web.loadbalancer.server.scheme=\"http\"", "--label", "traefik.http.routers.app-web.rule=\"PathPrefix(\\`/\\`)\"", "--label", "traefik.http.routers.app-web.priority=\"2\"", "--label", "traefik.http.middlewares.app-web-retry.retry.attempts=\"5\"", "--label", "traefik.http.middlewares.app-web-retry.retry.initialinterval=\"500ms\"", "--label", "traefik.http.routers.app-web.middlewares=\"app-web-retry@docker\"" ], @config.role(:web).label_args
   end
 
   test "custom labels" do
@@ -66,7 +66,7 @@ class ConfigurationRoleTest < ActiveSupport::TestCase
       c[:servers]["beta"] = { "traefik" => "true", "hosts" => [ "1.1.1.5" ] }
     })
 
-    assert_equal [ "--label", "service=\"app\"", "--label", "role=\"beta\"", "--label", "traefik.http.services.app-beta.loadbalancer.server.scheme=\"http\"", "--label", "traefik.http.routers.app-beta.rule=\"PathPrefix(\\`/\\`)\"", "--label", "traefik.http.middlewares.app-beta-retry.retry.attempts=\"5\"", "--label", "traefik.http.middlewares.app-beta-retry.retry.initialinterval=\"500ms\"", "--label", "traefik.http.routers.app-beta.middlewares=\"app-beta-retry@docker\"" ], config.role(:beta).label_args
+    assert_equal [ "--label", "service=\"app\"", "--label", "role=\"beta\"", "--label", "traefik.http.services.app-beta.loadbalancer.server.scheme=\"http\"", "--label", "traefik.http.routers.app-beta.rule=\"PathPrefix(\\`/\\`)\"", "--label", "traefik.http.routers.app-beta.priority=\"2\"", "--label", "traefik.http.middlewares.app-beta-retry.retry.attempts=\"5\"", "--label", "traefik.http.middlewares.app-beta-retry.retry.initialinterval=\"500ms\"", "--label", "traefik.http.routers.app-beta.middlewares=\"app-beta-retry@docker\"" ], config.role(:beta).label_args
   end
 
   test "env overwritten by role" do
@@ -176,6 +176,34 @@ class ConfigurationRoleTest < ActiveSupport::TestCase
     ENV["REDIS_PASSWORD"] = nil
   end
 
+  test "env overwritten by role with secrets" do
+    @deploy_with_roles[:env] = {
+      "clear" => {
+        "REDIS_URL" => "redis://a/b"
+      },
+      "secret" => [
+        "REDIS_PASSWORD"
+      ]
+    }
+
+    @deploy_with_roles[:servers]["workers"]["env"] = {
+      "clear" => {
+        "REDIS_URL" => "redis://c/d",
+      },
+    }
+
+    ENV["REDIS_PASSWORD"] = "secret456"
+
+    expected = <<~ENV
+      REDIS_PASSWORD=secret456
+      REDIS_URL=redis://c/d
+    ENV
+
+    assert_equal expected, @config_with_roles.role(:workers).env_file.to_s
+  ensure
+    ENV["REDIS_PASSWORD"] = nil
+  end
+
   test "host_env_directory" do
     assert_equal ".kamal/env/roles", @config_with_roles.role(:workers).host_env_directory
   end

test/configuration/ssh_test.rb

@@ -22,6 +22,9 @@ class ConfigurationSshTest < ActiveSupport::TestCase
 
     config = Kamal::Configuration.new(@deploy.tap { |c| c.merge!(ssh: { "log_level" => "debug" }) })
     assert_equal 0, config.ssh.options[:logger].level
+
+    config = Kamal::Configuration.new(@deploy.tap { |c| c.merge!(ssh: { "port" => 2222 }) })
+    assert_equal 2222, config.ssh.options[:port]
   end
 
   test "ssh options with proxy host" do

test/configuration_test.rb

@@ -42,6 +42,16 @@ class ConfigurationTest < ActiveSupport::TestCase
     end
   end
 
+  test "service name valid" do
+    assert Kamal::Configuration.new(@deploy.tap { _1[:service] = "hey-app1_primary" }).valid?
+  end
+
+  test "service name invalid" do
+    assert_raise(ArgumentError) do
+      Kamal::Configuration.new @deploy.tap { _1[:service] = "app.com" }
+    end
+  end
+
   test "roles" do
     assert_equal %w[ web ], @config.roles.collect(&:name)
     assert_equal %w[ web workers ], @config_with_roles.roles.collect(&:name)
@@ -58,9 +68,9 @@ class ConfigurationTest < ActiveSupport::TestCase
     assert_equal [ "1.1.1.1", "1.1.1.2", "1.1.1.3" ], @config_with_roles.all_hosts
   end
 
-  test "primary web host" do
-    assert_equal "1.1.1.1", @config.primary_web_host
-    assert_equal "1.1.1.1", @config_with_roles.primary_web_host
+  test "primary host" do
+    assert_equal "1.1.1.1", @config.primary_host
+    assert_equal "1.1.1.1", @config_with_roles.primary_host
   end
 
   test "traefik hosts" do
@@ -128,14 +138,6 @@ class ConfigurationTest < ActiveSupport::TestCase
     assert_equal "healthcheck-app", @config.healthcheck_service
   end
 
-  test "env with missing secret" do
-    assert_raises(KeyError) do
-      config = Kamal::Configuration.new(@deploy.tap { |c| c.merge!({
-        env: { "secret" => [ "PASSWORD" ] }
-      }) }).ensure_env_available
-    end
-  end
-
   test "valid config" do
     assert @config.valid?
     assert @config_with_roles.valid?
@@ -173,6 +175,16 @@ class ConfigurationTest < ActiveSupport::TestCase
     end
   end
 
+  test "allow_empty_roles" do
+    assert_silent do
+      Kamal::Configuration.new @deploy.merge(servers: { "web" => %w[ web ], "workers" => { "hosts" => %w[ ] } }, allow_empty_roles: true)
+    end
+
+    assert_raises(ArgumentError) do
+      Kamal::Configuration.new @deploy.merge(servers: { "web" => %w[], "workers" => { "hosts" => %w[] } }, allow_empty_roles: true)
+    end
+  end
+
   test "volume_args" do
     assert_equal ["--volume", "/local/path:/container/path"], @config.volume_args
   end
@@ -235,7 +247,7 @@ class ConfigurationTest < ActiveSupport::TestCase
         :repository=>"dhh/app",
         :absolute_image=>"dhh/app:missing",
         :service_with_version=>"app-missing",
-        :ssh_options=>{ :user=>"root", log_level: :fatal, keepalive: true, keepalive_interval: 30 },
+        :ssh_options=>{ :user=>"root", port: 22, log_level: :fatal, keepalive: true, keepalive_interval: 30 },
         :sshkit=>{},
         :volume_args=>["--volume", "/local/path:/container/path"],
         :builder=>{},
@@ -286,4 +298,33 @@ class ConfigurationTest < ActiveSupport::TestCase
     assert_nil @config.asset_path
     assert_equal "foo", Kamal::Configuration.new(@deploy.merge!(asset_path: "foo")).asset_path
   end
+
+  test "primary role" do
+    assert_equal "web", @config.primary_role.name
+
+    config = Kamal::Configuration.new(@deploy_with_roles.deep_merge({
+      servers: { "alternate_web" => { "hosts" => [ "1.1.1.4", "1.1.1.5" ] } },
+      primary_role: "alternate_web" } ))
+
+
+    assert_equal "alternate_web", config.primary_role.name
+    assert_equal "1.1.1.4", config.primary_host
+    assert config.role(:alternate_web).primary?
+    assert config.role(:alternate_web).running_traefik?
+  end
+
+  test "primary role missing" do
+    error = assert_raises(ArgumentError) do
+      Kamal::Configuration.new(@deploy.merge(primary_role: "bar"))
+    end
+    assert_match /bar isn't defined/, error.message
+  end
+
+  test "retain_containers" do
+    assert_equal 5, @config.retain_containers
+    config = Kamal::Configuration.new(@deploy_with_roles.merge(retain_containers: 2))
+    assert_equal 2, config.retain_containers
+
+    assert_raises(ArgumentError) { Kamal::Configuration.new(@deploy_with_roles.merge(retain_containers: 0)) }
+  end
 end

test/fixtures/deploy_primary_web_role_override.yml

@@ -0,0 +1,20 @@
+service: app
+image: dhh/app
+servers:
+  web_chicago:
+    traefik: enabled
+    hosts:
+      - 1.1.1.1
+      - 1.1.1.2
+  web_tokyo:
+    traefik: enabled
+    hosts:
+      - 1.1.1.3
+      - 1.1.1.4
+env:
+  REDIS_URL: redis://x/y
+registry:
+  server: registry.digitalocean.com
+  username: user
+  password: pw
+primary_role: web_tokyo

test/fixtures/deploy_with_aliases.yml

@@ -0,0 +1,36 @@
+# helper aliases
+chicago_hosts: &chicago_hosts
+  hosts:
+    - 1.1.1.1
+    - 1.1.1.2
+tokyo_hosts: &tokyo_hosts
+  hosts:
+    - 1.1.1.3
+    - 1.1.1.4
+web_common: &web_common
+  env:
+    ROLE: "web"
+  traefik: true
+
+# actual config
+service: app
+image: dhh/app
+servers:
+  web:
+    <<: *chicago_hosts
+    <<: *web_common
+  web_tokyo:
+    <<: *tokyo_hosts
+    <<: *web_common
+  workers:
+    cmd: bin/jobs
+    <<: *chicago_hosts
+  workers_tokyo:
+    cmd: bin/jobs
+    <<: *tokyo_hosts
+env:
+  REDIS_URL: redis://x/y
+registry:
+  server: registry.digitalocean.com
+  username: user
+  password: pw

test/fixtures/deploy_with_low_percentage_boot_strategy.yml

@@ -0,0 +1,17 @@
+service: app
+image: dhh/app
+servers:
+  web:
+    - "1.1.1.1"
+    - "1.1.1.2"
+  workers:
+    - "1.1.1.3"
+    - "1.1.1.4"
+
+registry:
+  username: user
+  password: pw
+
+boot:
+  limit: 1%
+  wait: 2

test/fixtures/deploy_with_percentage_boot_strategy.yml

@@ -13,5 +13,5 @@ registry:
   password: pw
 
 boot:
-  limit: 25%
+  limit: 1%
   wait: 2

test/fixtures/deploy_workers_only.yml

@@ -0,0 +1,12 @@
+service: app
+image: dhh/app
+servers:
+  workers:
+    traefik: false
+    hosts:
+      - 1.1.1.1
+      - 1.1.1.2
+primary_role: workers
+registry:
+  username: user
+  password: pw

test/integration/app_test.rb

@@ -10,8 +10,7 @@ class AppTest < IntegrationTest
 
     kamal :app, :stop
 
-    # traefik is up and returns 404s when it can't match a route
-    assert_app_not_found
+    assert_app_is_down
 
     kamal :app, :start
 
@@ -51,7 +50,6 @@ class AppTest < IntegrationTest
 
     kamal :app, :remove
 
-    # traefik is up and returns 404s when it can't match a route
-    assert_app_not_found
+    assert_app_is_down
   end
 end

test/integration/docker/deployer/app/.kamal/hooks/docker-setup

@@ -0,0 +1,3 @@
+#!/bin/sh
+echo "Docker set up!"
+mkdir -p /tmp/${TEST_ID} && touch /tmp/${TEST_ID}/docker-setup

test/integration/docker/deployer/app/.kamal/hooks/post-traefik-reboot

@@ -0,0 +1,3 @@
+#!/bin/sh
+echo "Rebooted Traefik on ${KAMAL_HOSTS}"
+mkdir -p /tmp/${TEST_ID} && touch /tmp/${TEST_ID}/post-traefik-reboot

test/integration/docker/deployer/app/.kamal/hooks/pre-traefik-reboot

@@ -0,0 +1,3 @@
+#!/bin/sh
+echo "Rebooting Traefik on ${KAMAL_HOSTS}..."
+mkdir -p /tmp/${TEST_ID} && touch /tmp/${TEST_ID}/pre-traefik-reboot

test/integration/docker/deployer/app/config/deploy.yml

@@ -24,9 +24,10 @@ traefik:
   args:
     accesslog: true
     accesslog.format: json
-  image: registry:4443/traefik:v2.9
+  image: registry:4443/traefik:v2.10
 accessories:
   busybox:
+    service: custom-busybox
     image: registry:4443/busybox:1.36.0
     cmd: sh -c 'echo "Starting busybox..."; trap exit term; while true; do sleep 1; done'
     roles:

test/integration/docker/deployer/setup.sh

@@ -19,5 +19,9 @@ push_image_to_registry_4443() {
 
 install_kamal
 push_image_to_registry_4443 nginx 1-alpine-slim
-push_image_to_registry_4443 traefik v2.9
+push_image_to_registry_4443 traefik v2.10
 push_image_to_registry_4443 busybox 1.36.0
+
+# .ssh is on a shared volume that persists between runs. Clean it up as the
+# churn of temporary vm IPs can eventually create conflicts.
+rm -f /root/.ssh/known_hosts

test/integration/integration_test.rb

@@ -55,12 +55,6 @@ class IntegrationTest < ActiveSupport::TestCase
       assert_app_version(version, response) if version
     end
 
-    def assert_app_not_found
-      response = app_response
-      debug_response_code(response, "404")
-      assert_equal "404", response.code
-    end
-
     def wait_for_app_to_be_up(timeout: 20, up_count: 3)
       timeout_at = Time.now + timeout
       up_times = 0
@@ -109,7 +103,7 @@ class IntegrationTest < ActiveSupport::TestCase
       assert_equal "200", code
     end
 
-    def wait_for_healthy(timeout: 20)
+    def wait_for_healthy(timeout: 30)
       timeout_at = Time.now + timeout
       while docker_compose("ps -a | tail -n +2 | grep -v '(healthy)' | wc -l", capture: true) != "0"
         if timeout_at < Time.now

test/integration/lock_test.rb

@@ -10,7 +10,7 @@ class LockTest < IntegrationTest
     assert_match /Locked by: Deployer at .*\nVersion: #{latest_app_version}\nMessage: Integration Tests/m, status
 
     error = kamal :deploy, capture: true, raise_on_error: false
-    assert_match /Deploy lock found/m, error
+    assert_match /Deploy lock found. Run 'kamal lock help' for more information/m, error
 
     kamal :lock, :release
 

test/integration/main_test.rb

@@ -5,6 +5,7 @@ class MainTest < IntegrationTest
     kamal :envify
     assert_local_env_file "SECRET_TOKEN=1234"
     assert_remote_env_file "SECRET_TOKEN=1234\nCLEAR_TOKEN=4321"
+    remove_local_env_file
 
     first_version = latest_app_version
 
@@ -31,7 +32,7 @@ class MainTest < IntegrationTest
     assert_match /Traefik Host: vm2/, details
     assert_match /App Host: vm1/, details
     assert_match /App Host: vm2/, details
-    assert_match /traefik:v2.9/, details
+    assert_match /traefik:v2.10/, details
     assert_match /registry:4443\/app:#{first_version}/, details
 
     audit = kamal :audit, capture: true
@@ -53,17 +54,34 @@ class MainTest < IntegrationTest
     assert_equal "registry:4443/app:#{version}", config[:absolute_image]
     assert_equal "app-#{version}", config[:service_with_version]
     assert_equal [], config[:volume_args]
-    assert_equal({ user: "root", keepalive: true, keepalive_interval: 30, log_level: :fatal }, config[:ssh_options])
+    assert_equal({ user: "root", port: 22, keepalive: true, keepalive_interval: 30, log_level: :fatal }, config[:ssh_options])
     assert_equal({ "multiarch" => false, "args" => { "COMMIT_SHA" => version } }, config[:builder])
     assert_equal [ "--log-opt", "max-size=\"10m\"" ], config[:logging]
     assert_equal({ "path" => "/up", "port" => 3000, "max_attempts" => 7, "exposed_port" => 3999, "cord"=>"/tmp/kamal-cord", "log_lines" => 50, "cmd"=>"wget -qO- http://localhost > /dev/null || exit 1" }, config[:healthcheck])
   end
 
+  test "setup and remove" do
+    # Check remove completes when nothing has been setup yet
+    kamal :remove, "-y"
+    assert_no_images_or_containers
+
+    kamal :envify
+    kamal :setup
+    assert_images_and_containers
+
+    kamal :remove, "-y"
+    assert_no_images_or_containers
+  end
+
   private
     def assert_local_env_file(contents)
       assert_equal contents, deployer_exec("cat .env", capture: true)
     end
 
+    def remove_local_env_file
+      deployer_exec("rm .env")
+    end
+
     def assert_remote_env_file(contents)
       assert_equal contents, docker_compose("exec vm1 cat /root/.kamal/env/roles/app-web.env", capture: true)
     end
@@ -79,4 +97,22 @@ class MainTest < IntegrationTest
 
       assert_equal "200", Net::HTTP.get_response(URI.parse("http://localhost:12345/versions/.hidden")).code
     end
+
+    def vm1_image_ids
+      docker_compose("exec vm1 docker image ls -q", capture: true).strip.split("\n")
+    end
+
+    def vm1_container_ids
+      docker_compose("exec vm1 docker ps -a -q", capture: true).strip.split("\n")
+    end
+
+    def assert_no_images_or_containers
+      assert vm1_image_ids.empty?
+      assert vm1_container_ids.empty?
+    end
+
+    def assert_images_and_containers
+      assert vm1_image_ids.any?
+      assert vm1_container_ids.any?
+    end
 end

test/integration/traefik_test.rb

@@ -7,8 +7,19 @@ class TraefikTest < IntegrationTest
     kamal :traefik, :boot
     assert_traefik_running
 
-    kamal :traefik, :reboot
+    output = kamal :traefik, :reboot, capture: true
     assert_traefik_running
+    assert_hooks_ran "pre-traefik-reboot", "post-traefik-reboot"
+    assert_match /Rebooting Traefik on vm1,vm2.../, output
+    assert_match /Rebooted Traefik on vm1,vm2/, output
+
+    output = kamal :traefik, :reboot, :"--rolling", capture: true
+    assert_traefik_running
+    assert_hooks_ran "pre-traefik-reboot", "post-traefik-reboot"
+    assert_match /Rebooting Traefik on vm1.../, output
+    assert_match /Rebooted Traefik on vm1/, output
+    assert_match /Rebooting Traefik on vm2.../, output
+    assert_match /Rebooted Traefik on vm2/, output
 
     kamal :traefik, :boot
     assert_traefik_running
@@ -41,11 +52,11 @@ class TraefikTest < IntegrationTest
 
   private
     def assert_traefik_running
-      assert_match /traefik:v2.9   "\/entrypoint.sh/, traefik_details
+      assert_match /traefik:v2.10   "\/entrypoint.sh/, traefik_details
     end
 
     def assert_traefik_not_running
-      refute_match /traefik:v2.9   "\/entrypoint.sh/, traefik_details
+      refute_match /traefik:v2.10   "\/entrypoint.sh/, traefik_details
     end
 
     def traefik_details