All count as proxy hosts

This will allow us to share files with the proxy via the host.

Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    kamal (2.0.0.rc2)
+    kamal (2.0.0.alpha)
       activesupport (>= 7.0)
       base64 (~> 0.2)
       bcrypt_pbkdf (~> 1.0)
@@ -9,7 +9,6 @@ PATH
       dotenv (~> 3.1)
       ed25519 (~> 1.2)
       net-ssh (~> 7.0)
-      net-ssh-gateway
       sshkit (>= 1.23.0, < 2.0)
       thor (~> 1.3)
       zeitwerk (~> 2.5)
@@ -80,8 +79,6 @@ GEM
     net-sftp (4.0.0)
       net-ssh (>= 5.0.0, < 8.0.0)
     net-ssh (7.2.3)
-    net-ssh-gateway (2.0.0)
-      net-ssh (>= 4.0.0)
     nokogiri (1.16.7-arm64-darwin)
       racc (~> 1.4)
     nokogiri (1.16.7-x86_64-darwin)

README.md

@@ -1,6 +1,6 @@
 # Kamal: Deploy web apps anywhere
 
-From bare metal to cloud VMs, deploy web apps anywhere with zero downtime. Kamal uses [kamal-proxy](https://github.com/basecamp/kamal-proxy) to seamlessly switch requests between containers. Works seamlessly across multiple servers, using SSHKit to execute commands. Originally built for Rails apps, Kamal will work with any type of web app that can be containerized with Docker.
+From bare metal to cloud VMs, deploy web apps anywhere with zero downtime. Kamal has the dynamic reverse-proxy Traefik hold requests while a new app container is started and the old one is stopped. Works seamlessly across multiple hosts, using SSHKit to execute commands. Originally built for Rails apps, Kamal will work with any type of web app that can be containerized with Docker.
 
 ➡️ See [kamal-deploy.org](https://kamal-deploy.org) for documentation on [installation](https://kamal-deploy.org/docs/installation), [configuration](https://kamal-deploy.org/docs/configuration), and [commands](https://kamal-deploy.org/docs/commands).
 

bin/docs

@@ -22,13 +22,15 @@ DOCS = {
   "builder" => "Builders",
   "configuration" => "Configuration overview",
   "env" => "Environment variables",
+  "healthcheck" => "Healthchecks",
   "logging" => "Logging",
-  "proxy" => "Proxy",
+  "proxy" => "Proxy (Experimental)",
   "registry" => "Docker Registry",
   "role" => "Roles",
   "servers" => "Servers",
   "ssh" => "SSH",
-  "sshkit" => "SSHKit"
+  "sshkit" => "SSHKit",
+  "traefik" => "Traefik"
 }
 
 class DocWriter

kamal.gemspec

@@ -14,7 +14,6 @@ Gem::Specification.new do |spec|
   spec.add_dependency "activesupport", ">= 7.0"
   spec.add_dependency "sshkit", ">= 1.23.0", "< 2.0"
   spec.add_dependency "net-ssh", "~> 7.0"
-  spec.add_dependency "net-ssh-gateway"
   spec.add_dependency "thor", "~> 1.3"
   spec.add_dependency "dotenv", "~> 3.1"
   spec.add_dependency "zeitwerk", "~> 2.5"

lib/kamal/cli/accessory.rb

@@ -147,25 +147,23 @@ 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 :grep_options, aliases: "-o", desc: "Additional options supplied to grep"
   option :follow, aliases: "-f", desc: "Follow logs on primary server (or specific host set by --hosts)"
-  option :skip_timestamps, type: :boolean, aliases: "-T", desc: "Skip appending timestamps to logging output"
   def logs(name)
     with_accessory(name) do |accessory, hosts|
       grep = options[:grep]
       grep_options = options[:grep_options]
-      timestamps = !options[:skip_timestamps]
 
       if options[:follow]
         run_locally do
           info "Following logs on #{hosts}..."
-          info accessory.follow_logs(timestamps: timestamps, grep: grep, grep_options: grep_options)
-          exec accessory.follow_logs(timestamps: timestamps, grep: grep, grep_options: grep_options)
+          info accessory.follow_logs(grep: grep, grep_options: grep_options)
+          exec accessory.follow_logs(grep: grep, grep_options: grep_options)
         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(hosts) do
-          puts capture_with_info(*accessory.logs(timestamps: timestamps, since: since, lines: lines, grep: grep, grep_options: grep_options))
+          puts capture_with_info(*accessory.logs(since: since, lines: lines, grep: grep, grep_options: grep_options))
         end
       end
     end
@@ -220,25 +218,6 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
     end
   end
 
-  desc "upgrade", "Upgrade accessories from Kamal 1.x to 2.0 (restart them in 'kamal' network)"
-  option :rolling, type: :boolean, default: false, desc: "Upgrade one host at a time"
-  option :confirmed, aliases: "-y", type: :boolean, default: false, desc: "Proceed without confirmation question"
-  def upgrade(name)
-    confirming "This will restart all accessories" do
-      with_lock do
-        host_groups = options[:rolling] ? KAMAL.accessory_hosts : [ KAMAL.accessory_hosts ]
-        host_groups.each do |hosts|
-          host_list = Array(hosts).join(",")
-          KAMAL.with_specific_hosts(hosts) do
-            say "Upgrading #{name} accessories on #{host_list}...", :magenta
-            reboot name
-            say "Upgraded #{name} accessories on #{host_list}...", :magenta
-          end
-        end
-      end
-    end
-  end
-
   private
     def with_accessory(name)
       if KAMAL.config.accessory(name)
@@ -275,7 +254,7 @@ class Kamal::Cli::Accessory < Kamal::Cli::Base
     def prepare(name)
       with_accessory(name) do |accessory, hosts|
         on(hosts) do
-          execute *KAMAL.registry.login unless KAMAL.config.registry.local?
+          execute *KAMAL.registry.login
           execute *KAMAL.docker.create_network
         rescue SSHKit::Command::Failed => e
           raise unless e.message.include?("already exists")

lib/kamal/cli/app.rb

@@ -4,7 +4,7 @@ class Kamal::Cli::App < Kamal::Cli::Base
     with_lock do
       say "Get most recent version available as an image...", :magenta unless options[:version]
       using_version(version_or_latest) do |version|
-        say "Start container with version #{version} (or reboot if already running)...", :magenta
+        say "Start container with version #{version} using a #{KAMAL.config.readiness_delay}s readiness delay (or reboot if already running)...", :magenta
 
         # Assets are prepared in a separate step to ensure they are on all hosts before booting
         on(KAMAL.hosts) do
@@ -42,12 +42,12 @@ class Kamal::Cli::App < Kamal::Cli::Base
           execute *KAMAL.auditor.record("Started app version #{KAMAL.config.version}"), verbosity: :debug
           execute *app.start, raise_on_non_zero_exit: false
 
-          if role.running_proxy?
+          if role.running_traefik? && KAMAL.proxy_host?(host)
             version = capture_with_info(*app.current_running_version, raise_on_non_zero_exit: false).strip
             endpoint = capture_with_info(*app.container_id_for_version(version)).strip
             raise Kamal::Cli::BootError, "Failed to get endpoint for #{role} on #{host}, did the container boot?" if endpoint.empty?
 
-            execute *app.deploy(target: endpoint)
+            execute *KAMAL.proxy.deploy(role.container_prefix, target: endpoint)
           end
         end
       end
@@ -64,11 +64,11 @@ class Kamal::Cli::App < Kamal::Cli::Base
           app = KAMAL.app(role: role, host: host)
           execute *KAMAL.auditor.record("Stopped app", role: role), verbosity: :debug
 
-          if role.running_proxy?
+          if role.running_traefik? && KAMAL.proxy_host?(host)
             version = capture_with_info(*app.current_running_version, raise_on_non_zero_exit: false).strip
             endpoint = capture_with_info(*app.container_id_for_version(version)).strip
             if endpoint.present?
-              execute *app.remove(target: endpoint), raise_on_non_zero_exit: false
+              execute *KAMAL.proxy.remove(role.container_prefix, target: endpoint), raise_on_non_zero_exit: false
             end
           end
 
@@ -188,14 +188,12 @@ class Kamal::Cli::App < Kamal::Cli::Base
   option :grep, aliases: "-g", desc: "Show lines with grep match only (use this to fetch specific requests by id)"
   option :grep_options, aliases: "-o", desc: "Additional options supplied to grep"
   option :follow, aliases: "-f", desc: "Follow log on primary server (or specific host set by --hosts)"
-  option :skip_timestamps, type: :boolean, aliases: "-T", desc: "Skip appending timestamps to logging output"
   def logs
     # FIXME: Catch when app containers aren't running
 
     grep = options[:grep]
     grep_options = options[:grep_options]
     since = options[:since]
-    timestamps = !options[:skip_timestamps]
 
     if options[:follow]
       lines = options[:lines].presence || ((since || grep) ? nil : 10) # Default to 10 lines if since or grep isn't set
@@ -207,8 +205,8 @@ class Kamal::Cli::App < Kamal::Cli::Base
         role = KAMAL.roles_on(KAMAL.primary_host).first
 
         app = KAMAL.app(role: role, host: host)
-        info app.follow_logs(host: KAMAL.primary_host, timestamps: timestamps, lines: lines, grep: grep, grep_options: grep_options)
-        exec app.follow_logs(host: KAMAL.primary_host, timestamps: timestamps, lines: lines, grep: grep, grep_options: grep_options)
+        info app.follow_logs(host: KAMAL.primary_host, lines: lines, grep: grep, grep_options: grep_options)
+        exec app.follow_logs(host: KAMAL.primary_host, lines: lines, grep: grep, grep_options: grep_options)
       end
     else
       lines = options[:lines].presence || ((since || grep) ? nil : 100) # Default to 100 lines if since or grep isn't set
@@ -218,7 +216,7 @@ class Kamal::Cli::App < Kamal::Cli::Base
 
         roles.each do |role|
           begin
-            puts_by_host host, capture_with_info(*KAMAL.app(role: role, host: host).logs(timestamps: timestamps, since: since, lines: lines, grep: grep, grep_options: grep_options))
+            puts_by_host host, capture_with_info(*KAMAL.app(role: role, host: host).logs(since: since, lines: lines, grep: grep, grep_options: grep_options))
           rescue SSHKit::Command::Failed
             puts_by_host host, "Nothing found"
           end
@@ -233,7 +231,6 @@ class Kamal::Cli::App < Kamal::Cli::Base
       stop
       remove_containers
       remove_images
-      remove_app_directory
     end
   end
 
@@ -275,20 +272,6 @@ class Kamal::Cli::App < Kamal::Cli::Base
     end
   end
 
-  desc "remove_app_directory", "Remove the service directory from servers", hide: true
-  def remove_app_directory
-    with_lock do
-      on(KAMAL.hosts) do |host|
-        roles = KAMAL.roles_on(host)
-
-        roles.each do |role|
-          execute *KAMAL.auditor.record("Removed #{KAMAL.config.app_directory} on all servers", role: role), verbosity: :debug
-          execute *KAMAL.server.remove_app_directory, raise_on_non_zero_exit: false
-        end
-      end
-    end
-  end
-
   desc "version", "Show app version currently running on servers"
   def version
     on(KAMAL.hosts) do |host|

lib/kamal/cli/app/boot.rb

@@ -1,7 +1,7 @@
 class Kamal::Cli::App::Boot
   attr_reader :host, :role, :version, :barrier, :sshkit
   delegate :execute, :capture_with_info, :capture_with_pretty_json, :info, :error, :upload!, to: :sshkit
-  delegate :assets?, :running_proxy?, to: :role
+  delegate :uses_cord?, :assets?, :running_traefik?, to: :role
 
   def initialize(host, role, sshkit, version, barrier)
     @host = host
@@ -50,17 +50,20 @@ class Kamal::Cli::App::Boot
       execute *app.ensure_env_directory
       upload! role.secrets_io(host), role.secrets_path, mode: "0600"
 
-      execute *app.run(hostname: hostname)
-      if running_proxy?
-        endpoint = capture_with_info(*app.container_id_for_version(version)).strip
-        raise Kamal::Cli::BootError, "Failed to get endpoint for #{role} on #{host}, did the container boot?" if endpoint.empty?
-        execute *app.deploy(target: endpoint)
+      if proxy_host?
+        execute *app.run_for_proxy(hostname: hostname)
+        if running_traefik?
+          endpoint = capture_with_info(*app.container_id_for_version(version)).strip
+          raise Kamal::Cli::BootError, "Failed to get endpoint for #{role} on #{host}, did the container boot?" if endpoint.empty?
+          execute *KAMAL.proxy.deploy(role.container_prefix, target: endpoint)
+        else
+          Kamal::Cli::Healthcheck::Poller.wait_for_healthy(pause_after_ready: true) { capture_with_info(*app.status(version: version)) }
+        end
       else
+        execute *app.tie_cord(role.cord_host_file) if uses_cord?
+        execute *app.run(hostname: hostname)
         Kamal::Cli::Healthcheck::Poller.wait_for_healthy(pause_after_ready: true) { capture_with_info(*app.status(version: version)) }
       end
-    rescue => e
-      error "Failed to boot #{role} on #{host}"
-      raise e
     end
 
     def stop_new_version
@@ -68,7 +71,16 @@ class Kamal::Cli::App::Boot
     end
 
     def stop_old_version(version)
+      if uses_cord? && !proxy_host?
+        cord = capture_with_info(*app.cord(version: version), raise_on_non_zero_exit: false).strip
+        if cord.present?
+          execute *app.cut_cord(cord)
+          Kamal::Cli::Healthcheck::Poller.wait_for_unhealthy(pause_after_ready: true) { capture_with_info(*app.status(version: version)) }
+        end
+      end
+
       execute *app.stop(version: version), raise_on_non_zero_exit: false
+
       execute *app.clean_up_assets if assets?
     end
 
@@ -122,4 +134,8 @@ class Kamal::Cli::App::Boot
     def queuer?
       barrier && !barrier_role?
     end
+
+    def proxy_host?
+      KAMAL.proxy_host?(host)
+    end
 end

lib/kamal/cli/base.rb

@@ -101,7 +101,7 @@ module Kamal::Cli
       end
 
       def acquire_lock
-        ensure_run_directory
+        ensure_run_and_locks_directory
 
         raise_if_locked do
           say "Acquiring the deploy lock...", :magenta
@@ -135,10 +135,8 @@ module Kamal::Cli
           details = { hosts: KAMAL.hosts.join(","), command: command, subcommand: subcommand }
 
           say "Running the #{hook} hook...", :magenta
-          with_env KAMAL.hook.env(**details, **extra_details) do
-            run_locally do
-              execute *KAMAL.hook.run(hook)
-            end
+          run_locally do
+            execute *KAMAL.hook.run(hook, **details, **extra_details)
           rescue SSHKit::Command::Failed => e
             raise HookError.new("Hook `#{hook}` failed:\n#{e.message}")
           end
@@ -176,23 +174,14 @@ module Kamal::Cli
         instance_variable_get("@_invocations").first
       end
 
-      def reset_invocation(cli_class)
-        instance_variable_get("@_invocations")[cli_class].pop
-      end
-
-      def ensure_run_directory
+      def ensure_run_and_locks_directory
         on(KAMAL.hosts) do
           execute(*KAMAL.server.ensure_run_directory)
         end
-      end
 
-      def with_env(env)
-        current_env = ENV.to_h.dup
-        ENV.update(env)
-        yield
-      ensure
-        ENV.clear
-        ENV.update(current_env)
+        on(KAMAL.primary_host) do
+          execute(*KAMAL.lock.ensure_locks_directory)
+        end
       end
   end
 end

lib/kamal/cli/build.rb

@@ -1,5 +1,4 @@
 require "uri"
-require "net/ssh"
 
 class Kamal::Cli::Build < Kamal::Cli::Base
   class BuildError < StandardError; end
@@ -31,38 +30,34 @@ class Kamal::Cli::Build < Kamal::Cli::Base
       say "Building with uncommitted changes:\n #{uncommitted_changes}", :yellow
     end
 
-    with_env(KAMAL.config.builder.secrets) do
-      run_locally do
-        begin
-          execute *KAMAL.builder.inspect_builder
-        rescue SSHKit::Command::Failed => e
-          if e.message =~ /(context not found|no builder|no compatible builder|does not exist)/
-            warn "Missing compatible builder, so creating a new one first"
-            begin
-              cli.remove
-            rescue SSHKit::Command::Failed
-              raise unless e.message =~ /(context not found|no builder|does not exist)/
-            end
-            cli.create
-          else
-            raise
+    run_locally do
+      begin
+        execute *KAMAL.builder.inspect_builder
+      rescue SSHKit::Command::Failed => e
+        if e.message =~ /(context not found|no builder|no compatible builder|does not exist)/
+          warn "Missing compatible builder, so creating a new one first"
+          begin
+            cli.remove
+          rescue SSHKit::Command::Failed
+            raise unless e.message =~ /(context not found|no builder|does not exist)/
           end
+          cli.create
+        else
+          raise
         end
+      end
 
-        # Get the command here to ensure the Dir.chdir doesn't interfere with it
-        push = KAMAL.builder.push
+      # Get the command here to ensure the Dir.chdir doesn't interfere with it
+      push = KAMAL.builder.push
 
-        KAMAL.with_verbosity(:debug) do
-          Dir.chdir(KAMAL.config.builder.build_directory) { execute *push }
-        end
+      KAMAL.with_verbosity(:debug) do
+        Dir.chdir(KAMAL.config.builder.build_directory) { execute *push, env: KAMAL.config.builder.secrets }
       end
     end
   end
 
   desc "pull", "Pull app image from registry onto servers"
   def pull
-    tunnels = Kamal::Cli::Tunnel::RemotePorts.new(KAMAL.hosts, KAMAL.config.registry.local_port).tap(&:open) if KAMAL.config.registry.local?
-
     if (first_hosts = mirror_hosts).any?
       #  Pull on a single host per mirror first to seed them
       say "Pulling image on #{first_hosts.join(", ")} to seed the #{"mirror".pluralize(first_hosts.count)}...", :magenta
@@ -72,8 +67,6 @@ class Kamal::Cli::Build < Kamal::Cli::Base
     else
       pull_on_hosts(KAMAL.hosts)
     end
-  ensure
-    tunnels&.close
   end
 
   desc "create", "Create a build setup"
@@ -157,7 +150,7 @@ class Kamal::Cli::Build < Kamal::Cli::Base
     end
 
     def pull_on_hosts(hosts)
-      on(hosts) do |host|
+      on(hosts) do
         execute *KAMAL.auditor.record("Pulled image with version #{KAMAL.config.version}"), verbosity: :debug
         execute *KAMAL.builder.clean, raise_on_non_zero_exit: false
         execute *KAMAL.builder.pull

lib/kamal/cli/healthcheck/poller.rb

@@ -1,30 +1,26 @@
 module Kamal::Cli::Healthcheck::Poller
   extend self
 
-  def wait_for_healthy(role, &block)
+  TRAEFIK_UPDATE_DELAY = 5
+
+
+  def wait_for_healthy(pause_after_ready: false, &block)
     attempt = 1
-    timeout_at = Time.now + KAMAL.config.deploy_timeout
-    readiness_delay = KAMAL.config.readiness_delay
+    max_attempts = KAMAL.config.healthcheck.max_attempts
 
     begin
-      status = block.call
-
-      if status == "running"
-        # Wait for the readiness delay and confirm it is still running
-        if readiness_delay > 0
-          info "Container is running, waiting for readiness delay of #{readiness_delay} seconds"
-          sleep readiness_delay
-          status = block.call
-        end
-      end
-
-      unless %w[ running healthy ].include?(status)
-        raise Kamal::Cli::Healthcheck::Error, "container not ready after #{KAMAL.config.deploy_timeout} seconds (#{status})"
+      case status = block.call
+      when "healthy"
+        sleep TRAEFIK_UPDATE_DELAY if pause_after_ready
+      when "running" # No health check configured
+        sleep KAMAL.config.readiness_delay if pause_after_ready
+      else
+        raise Kamal::Cli::Healthcheck::Error, "container not ready (#{status})"
       end
     rescue Kamal::Cli::Healthcheck::Error => e
-      time_left = timeout_at - Time.now
-      if time_left > 0
-        sleep [ attempt, time_left ].min
+      if attempt <= max_attempts
+        info "#{e.message}, retrying in #{attempt}s (attempt #{attempt}/#{max_attempts})..."
+        sleep attempt
         attempt += 1
         retry
       else
@@ -35,6 +31,31 @@ module Kamal::Cli::Healthcheck::Poller
     info "Container is healthy!"
   end
 
+  def wait_for_unhealthy(pause_after_ready: false, &block)
+    attempt = 1
+    max_attempts = KAMAL.config.healthcheck.max_attempts
+
+    begin
+      case status = block.call
+      when "unhealthy"
+        sleep TRAEFIK_UPDATE_DELAY if pause_after_ready
+      else
+        raise Kamal::Cli::Healthcheck::Error, "container not unhealthy (#{status})"
+      end
+    rescue Kamal::Cli::Healthcheck::Error => e
+      if attempt <= max_attempts
+        info "#{e.message}, retrying in #{attempt}s (attempt #{attempt}/#{max_attempts})..."
+        sleep attempt
+        attempt += 1
+        retry
+      else
+        raise
+      end
+    end
+
+    info "Container is unhealthy!"
+  end
+
   private
     def info(message)
       SSHKit.config.output.info(message)

lib/kamal/cli/lock.rb

@@ -12,7 +12,7 @@ class Kamal::Cli::Lock < Kamal::Cli::Base
   option :message, aliases: "-m", type: :string, desc: "A lock message", required: true
   def acquire
     message = options[:message]
-    ensure_run_directory
+    ensure_run_and_locks_directory
 
     raise_if_locked do
       on(KAMAL.primary_host) do

lib/kamal/cli/main.rb

@@ -22,7 +22,7 @@ class Kamal::Cli::Main < Kamal::Cli::Base
       invoke_options = deploy_options
 
       say "Log into image registry...", :magenta
-      invoke "kamal:cli:registry:setup", [], invoke_options.merge(skip_local: options[:skip_push])
+      invoke "kamal:cli:registry:login", [], invoke_options.merge(skip_local: options[:skip_push])
 
       if options[:skip_push]
         say "Pull app image...", :magenta
@@ -35,8 +35,13 @@ class Kamal::Cli::Main < Kamal::Cli::Base
       with_lock do
         run_hook "pre-deploy", secrets: true
 
-        say "Ensure kamal-proxy is running...", :magenta
-        invoke "kamal:cli:proxy:boot", [], invoke_options
+        if KAMAL.config.proxy.enabled?
+          say "Ensure Traefik/kamal-proxy is running...", :magenta
+          invoke "kamal:cli:proxy:boot", [], invoke_options
+        else
+          say "Ensure Traefik is running...", :magenta
+          invoke "kamal:cli:traefik:boot", [], invoke_options
+        end
 
         say "Detect stale containers...", :magenta
         invoke "kamal:cli:app:stale_containers", [], invoke_options.merge(stop: true)
@@ -48,10 +53,10 @@ class Kamal::Cli::Main < Kamal::Cli::Base
       end
     end
 
-    run_hook "post-deploy", secrets: true, runtime: runtime.round.to_s
+    run_hook "post-deploy", secrets: true, runtime: runtime.round
   end
 
-  desc "redeploy", "Deploy app to servers without bootstrapping servers, starting kamal-proxy, pruning, and registry login"
+  desc "redeploy", "Deploy app to servers without bootstrapping servers, starting Traefik, pruning, and registry login"
   option :skip_push, aliases: "-P", type: :boolean, default: false, desc: "Skip image build and push"
   def redeploy
     runtime = print_runtime do
@@ -75,7 +80,7 @@ class Kamal::Cli::Main < Kamal::Cli::Base
       end
     end
 
-    run_hook "post-deploy", secrets: true, runtime: runtime.round.to_s
+    run_hook "post-deploy", secrets: true, runtime: runtime.round
   end
 
   desc "rollback [VERSION]", "Rollback app to VERSION"
@@ -99,12 +104,16 @@ class Kamal::Cli::Main < Kamal::Cli::Base
       end
     end
 
-    run_hook "post-deploy", secrets: true, runtime: runtime.round.to_s if rolled_back
+    run_hook "post-deploy", secrets: true, runtime: runtime.round if rolled_back
   end
 
   desc "details", "Show details about all containers"
   def details
-    invoke "kamal:cli:proxy:details"
+    if KAMAL.config.proxy.enabled?
+      invoke "kamal:cli:proxy:details"
+    else
+      invoke "kamal:cli:traefik:details"
+    end
     invoke "kamal:cli:app:details"
     invoke "kamal:cli:accessory:details", [ "all" ]
   end
@@ -123,7 +132,7 @@ class Kamal::Cli::Main < Kamal::Cli::Base
     end
   end
 
-  desc "docs [SECTION]", "Show Kamal configuration documentation"
+  desc "docs", "Show Kamal documentation for configuration setting"
   def docs(section = nil)
     case section
     when NilClass
@@ -176,46 +185,19 @@ class Kamal::Cli::Main < Kamal::Cli::Base
     end
   end
 
-  desc "remove", "Remove kamal-proxy, app, accessories, and registry session from servers"
+  desc "remove", "Remove Traefik, app, accessories, and registry session from servers"
   option :confirmed, aliases: "-y", type: :boolean, default: false, desc: "Proceed without confirmation question"
   def remove
     confirming "This will remove all containers and images. Are you sure?" do
       with_lock do
-        invoke "kamal:cli:app:remove", [], options.without(:confirmed)
-        invoke "kamal:cli:proxy:remove", [], options.without(:confirmed)
-        invoke "kamal:cli:accessory:remove", [ "all" ], options
-        invoke "kamal:cli:registry:remove", [], options.without(:confirmed).merge(skip_local: true)
-      end
-    end
-  end
-
-  desc "upgrade", "Upgrade from Kamal 1.x to 2.0"
-  option :confirmed, aliases: "-y", type: :boolean, default: false, desc: "Proceed without confirmation question"
-  option :rolling, type: :boolean, default: false, desc: "Upgrade one host at a time"
-  def upgrade
-    confirming "This will replace Traefik with kamal-proxy and restart all accessories" do
-      with_lock do
-        if options[:rolling]
-          (KAMAL.hosts | KAMAL.accessory_hosts).each do |host|
-            KAMAL.with_specific_hosts(host) do
-              say "Upgrading #{host}...", :magenta
-              if KAMAL.hosts.include?(host)
-                invoke "kamal:cli:proxy:upgrade", [], options.merge(confirmed: true, rolling: false)
-                reset_invocation(Kamal::Cli::Proxy)
-              end
-              if KAMAL.accessory_hosts.include?(host)
-                invoke "kamal:cli:accessory:upgrade", [ "all" ], options.merge(confirmed: true, rolling: false)
-                reset_invocation(Kamal::Cli::Accessory)
-              end
-              say "Upgraded #{host}", :magenta
-            end
-          end
+        if KAMAL.config.proxy.enabled?
+          invoke "kamal:cli:proxy:remove", [], options.without(:confirmed)
         else
-          say "Upgrading all hosts...", :magenta
-          invoke "kamal:cli:proxy:upgrade", [], options.merge(confirmed: true)
-          invoke "kamal:cli:accessory:upgrade", [ "all" ], options.merge(confirmed: true)
-          say "Upgraded all hosts", :magenta
+          invoke "kamal:cli:traefik:remove", [], options.without(:confirmed)
         end
+        invoke "kamal:cli:app:remove", [], options.without(:confirmed)
+        invoke "kamal:cli:accessory:remove", [ "all" ], options
+        invoke "kamal:cli:registry:logout", [], options.without(:confirmed).merge(skip_local: true)
       end
     end
   end
@@ -237,7 +219,7 @@ class Kamal::Cli::Main < Kamal::Cli::Base
   desc "lock", "Manage the deploy lock"
   subcommand "lock", Kamal::Cli::Lock
 
-  desc "proxy", "Manage kamal-proxy"
+  desc "proxy", "Prune old application images and containers"
   subcommand "proxy", Kamal::Cli::Proxy
 
   desc "prune", "Prune old application images and containers"
@@ -252,6 +234,9 @@ class Kamal::Cli::Main < Kamal::Cli::Base
   desc "server", "Bootstrap servers with curl and Docker"
   subcommand "server", Kamal::Cli::Server
 
+  desc "traefik", "Manage Traefik load balancer"
+  subcommand "traefik", Kamal::Cli::Traefik
+
   private
     def container_available?(version)
       begin

lib/kamal/cli/proxy.rb

@@ -1,6 +1,7 @@
 class Kamal::Cli::Proxy < Kamal::Cli::Base
   desc "boot", "Boot proxy on servers"
   def boot
+    raise_unless_kamal_proxy_enabled!
     with_lock do
       on(KAMAL.hosts) do |host|
         execute *KAMAL.docker.create_network
@@ -8,15 +9,15 @@ class Kamal::Cli::Proxy < Kamal::Cli::Base
         raise unless e.message.include?("already exists")
       end
 
-      on(KAMAL.proxy_hosts) do |host|
-        execute *KAMAL.registry.login unless KAMAL.config.registry.local?
-
-        version = capture_with_info(*KAMAL.proxy.version).strip.presence
-
-        if version && Kamal::Utils.older_version?(version, Kamal::Configuration::PROXY_MINIMUM_VERSION)
-          raise "kamal-proxy version #{version} is too old, please reboot to update to at least #{Kamal::Configuration::PROXY_MINIMUM_VERSION}"
+      on(KAMAL.traefik_hosts) do |host|
+        execute *KAMAL.registry.login
+        if KAMAL.proxy_host?(host)
+          execute *KAMAL.proxy.start_or_run
+        else
+          execute *KAMAL.traefik.ensure_env_directory
+          upload! KAMAL.traefik.secrets_io, KAMAL.traefik.secrets_path, mode: "0600"
+          execute *KAMAL.traefik.start_or_run
         end
-        execute *KAMAL.proxy.start_or_run
       end
     end
   end
@@ -25,105 +26,109 @@ class Kamal::Cli::Proxy < Kamal::Cli::Base
   option :rolling, type: :boolean, default: false, desc: "Reboot proxy on hosts in sequence, rather than in parallel"
   option :confirmed, aliases: "-y", type: :boolean, default: false, desc: "Proceed without confirmation question"
   def reboot
+    raise_unless_kamal_proxy_enabled!
     confirming "This will cause a brief outage on each host. Are you sure?" do
       with_lock do
-        host_groups = options[:rolling] ? KAMAL.proxy_hosts : [ KAMAL.proxy_hosts ]
+        host_groups = options[:rolling] ? KAMAL.traefik_hosts : [ KAMAL.traefik_hosts ]
         host_groups.each do |hosts|
           host_list = Array(hosts).join(",")
-          run_hook "pre-proxy-reboot", hosts: host_list
+          run_hook "pre-traefik-reboot", hosts: host_list
           on(hosts) do |host|
             execute *KAMAL.auditor.record("Rebooted proxy"), verbosity: :debug
-            execute *KAMAL.registry.login unless KAMAL.config.registry.local?
+            execute *KAMAL.registry.login
 
             "Stopping and removing Traefik on #{host}, if running..."
-            execute *KAMAL.proxy.cleanup_traefik
+            execute *KAMAL.traefik.stop, raise_on_non_zero_exit: false
+            execute *KAMAL.traefik.remove_container
 
             "Stopping and removing kamal-proxy on #{host}, if running..."
             execute *KAMAL.proxy.stop, raise_on_non_zero_exit: false
             execute *KAMAL.proxy.remove_container
 
-            execute *KAMAL.proxy.run
+            execute *KAMAL.traefik_or_proxy(host).run
 
-            KAMAL.roles_on(host).select(&:running_proxy?).each do |role|
-              app = KAMAL.app(role: role, host: host)
+            if KAMAL.proxy_host?(host)
+              KAMAL.roles_on(host).select(&:running_traefik?).each do |role|
+                app = KAMAL.app(role: role, host: host)
 
-              version = capture_with_info(*app.current_running_version, raise_on_non_zero_exit: false).strip
-              endpoint = capture_with_info(*app.container_id_for_version(version)).strip
+                version = capture_with_info(*app.current_running_version, raise_on_non_zero_exit: false).strip
+                endpoint = capture_with_info(*app.container_id_for_version(version)).strip
 
-              if endpoint.present?
-                info "Deploying #{endpoint} for role `#{role}` on #{host}..."
-                execute *app.deploy(target: endpoint)
+                if endpoint.present?
+                  info "Deploying #{endpoint} for role `#{role}` on #{host}..."
+                  execute *KAMAL.proxy.deploy(role.container_prefix, target: endpoint)
+                end
               end
             end
           end
-          run_hook "post-proxy-reboot", hosts: host_list
+          run_hook "post-traefik-reboot", hosts: host_list
         end
       end
     end
   end
 
-  desc "upgrade", "Upgrade to kamal-proxy on servers (stop container, remove container, start new container, reboot app)", hide: true
+  desc "upgrade", "Upgrade to correct proxy on servers (stop container, remove container, start new container)"
   option :rolling, type: :boolean, default: false, desc: "Reboot proxy on hosts in sequence, rather than in parallel"
   option :confirmed, aliases: "-y", type: :boolean, default: false, desc: "Proceed without confirmation question"
   def upgrade
-    invoke_options = { "version" => KAMAL.config.latest_tag }.merge(options)
+    invoke_options = { "version" => KAMAL.config.version }.merge(options)
 
+    raise_unless_kamal_proxy_enabled!
     confirming "This will cause a brief outage on each host. Are you sure?" do
       host_groups = options[:rolling] ? KAMAL.hosts : [ KAMAL.hosts ]
       host_groups.each do |hosts|
         host_list = Array(hosts).join(",")
-        say "Upgrading proxy on #{host_list}...", :magenta
-        run_hook "pre-proxy-reboot", hosts: host_list
+        run_hook "pre-traefik-reboot", hosts: host_list
         on(hosts) do |host|
           execute *KAMAL.auditor.record("Rebooted proxy"), verbosity: :debug
-          execute *KAMAL.registry.login unless KAMAL.config.registry.local?
+          execute *KAMAL.registry.login
 
           "Stopping and removing Traefik on #{host}, if running..."
-          execute *KAMAL.proxy.cleanup_traefik
+          execute *KAMAL.traefik.stop, raise_on_non_zero_exit: false
+          execute *KAMAL.traefik.remove_container
 
           "Stopping and removing kamal-proxy on #{host}, if running..."
           execute *KAMAL.proxy.stop, raise_on_non_zero_exit: false
           execute *KAMAL.proxy.remove_container
-          execute *KAMAL.proxy.remove_image
         end
 
-        KAMAL.with_specific_hosts(hosts) do
-          invoke "kamal:cli:proxy:boot", [], invoke_options
-          reset_invocation(Kamal::Cli::Proxy)
-          invoke "kamal:cli:app:boot", [], invoke_options
-          reset_invocation(Kamal::Cli::App)
-          invoke "kamal:cli:prune:all", [], invoke_options
-          reset_invocation(Kamal::Cli::Prune)
-        end
+        invoke "kamal:cli:proxy:boot", [], invoke_options.merge("hosts" => host_list)
+        reset_invocation(Kamal::Cli::Proxy)
+        invoke "kamal:cli:app:boot", [], invoke_options.merge("hosts" => host_list, version: KAMAL.config.latest_tag)
+        reset_invocation(Kamal::Cli::App)
+        invoke "kamal:cli:prune:all", [], invoke_options.merge("hosts" => host_list)
+        reset_invocation(Kamal::Cli::Prune)
 
-        run_hook "post-proxy-reboot", hosts: host_list
-        say "Upgraded proxy on #{host_list}", :magenta
+        run_hook "post-traefik-reboot", hosts: host_list
       end
     end
   end
 
   desc "start", "Start existing proxy container on servers"
   def start
+    raise_unless_kamal_proxy_enabled!
     with_lock do
-      on(KAMAL.proxy_hosts) do |host|
+      on(KAMAL.traefik_hosts) do |host|
         execute *KAMAL.auditor.record("Started proxy"), verbosity: :debug
-        execute *KAMAL.proxy.start
+        execute *KAMAL.traefik_or_proxy(host).start
       end
     end
   end
 
   desc "stop", "Stop existing proxy container on servers"
   def stop
+    raise_unless_kamal_proxy_enabled!
     with_lock do
-      on(KAMAL.proxy_hosts) do |host|
+      on(KAMAL.traefik_hosts) do |host|
         execute *KAMAL.auditor.record("Stopped proxy"), verbosity: :debug
-        execute *KAMAL.proxy.stop, raise_on_non_zero_exit: false
+        execute *KAMAL.traefik_or_proxy(host).stop, raise_on_non_zero_exit: false
       end
     end
   end
 
   desc "restart", "Restart existing proxy container on servers"
   def restart
+    raise_unless_kamal_proxy_enabled!
     with_lock do
       stop
       start
@@ -132,7 +137,8 @@ class Kamal::Cli::Proxy < Kamal::Cli::Base
 
   desc "details", "Show details about proxy container from servers"
   def details
-    on(KAMAL.proxy_hosts) { |host| puts_by_host host, capture_with_info(*KAMAL.proxy.info), type: "Proxy" }
+    raise_unless_kamal_proxy_enabled!
+    on(KAMAL.traefik_hosts) { |host| puts_by_host host, capture_with_info(*KAMAL.traefik_or_proxy(host).info), type: "Proxy" }
   end
 
   desc "logs", "Show log lines from proxy on servers"
@@ -140,76 +146,68 @@ class Kamal::Cli::Proxy < Kamal::Cli::Base
   option :lines, type: :numeric, aliases: "-n", desc: "Number of log lines to pull from each server"
   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)"
-  option :skip_timestamps, type: :boolean, aliases: "-T", desc: "Skip appending timestamps to logging output"
   def logs
+    raise_unless_kamal_proxy_enabled!
     grep = options[:grep]
-    timestamps = !options[:skip_timestamps]
 
     if options[:follow]
       run_locally do
         info "Following logs on #{KAMAL.primary_host}..."
-        info KAMAL.proxy.follow_logs(host: KAMAL.primary_host, timestamps: timestamps, grep: grep)
-        exec KAMAL.proxy.follow_logs(host: KAMAL.primary_host, timestamps: timestamps, grep: grep)
+        info KAMAL.traefik_or_proxy(KAMAL.primary_host).follow_logs(host: KAMAL.primary_host, grep: grep)
+        exec KAMAL.traefik_or_proxy(KAMAL.primary_host).follow_logs(host: KAMAL.primary_host, 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.proxy_hosts) do |host|
-        puts_by_host host, capture(*KAMAL.proxy.logs(timestamps: timestamps, since: since, lines: lines, grep: grep)), type: "Proxy"
+      on(KAMAL.traefik_hosts) do |host|
+        puts_by_host host, capture(*KAMAL.traefik_or_proxy(host).logs(since: since, lines: lines, grep: grep)), type: "Proxy"
       end
     end
   end
 
   desc "remove", "Remove proxy container and image from servers"
-  option :force, type: :boolean, default: false, desc: "Force removing proxy when apps are still installed"
   def remove
+    raise_unless_kamal_proxy_enabled!
     with_lock do
-      if removal_allowed?(options[:force])
-        stop
-        remove_container
-        remove_image
-      end
+      stop
+      remove_container
+      remove_image
     end
   end
 
   desc "remove_container", "Remove proxy container from servers", hide: true
   def remove_container
+    raise_unless_kamal_proxy_enabled!
     with_lock do
-      on(KAMAL.proxy_hosts) do
+      on(KAMAL.traefik_hosts) do
         execute *KAMAL.auditor.record("Removed proxy container"), verbosity: :debug
         execute *KAMAL.proxy.remove_container
+        execute *KAMAL.traefik.remove_container
       end
     end
   end
 
   desc "remove_image", "Remove proxy image from servers", hide: true
   def remove_image
+    raise_unless_kamal_proxy_enabled!
     with_lock do
-      on(KAMAL.proxy_hosts) do
+      on(KAMAL.traefik_hosts) do
         execute *KAMAL.auditor.record("Removed proxy image"), verbosity: :debug
         execute *KAMAL.proxy.remove_image
+        execute *KAMAL.traefik.remove_image
       end
     end
   end
 
   private
-    def removal_allowed?(force)
-      on(KAMAL.proxy_hosts) do |host|
-        app_count = capture_with_info(*KAMAL.server.app_directory_count).chomp.to_i
-        raise "The are other applications installed on #{host}" if app_count > 0
+    def raise_unless_kamal_proxy_enabled!
+      unless KAMAL.config.proxy.enabled?
+        raise "kamal proxy commands are disabled unless experimental proxy support is enabled. Use `kamal traefik` commands instead."
       end
+    end
 
-      true
-    rescue SSHKit::Runner::ExecuteError => e
-      raise unless e.message.include?("The are other applications installed on")
-
-      if force
-        say "Forcing, so removing the proxy, even though other apps are installed", :magenta
-      else
-        say "Not removing the proxy, as other apps are installed, ignore this check with kamal proxy remove --force", :magenta
-      end
-
-      force
+    def reset_invocation(cli_class)
+      instance_variable_get("@_invocations")[cli_class].pop
     end
 end

lib/kamal/cli/prune.rb

@@ -28,6 +28,7 @@ class Kamal::Cli::Prune < Kamal::Cli::Base
       on(KAMAL.hosts) do
         execute *KAMAL.auditor.record("Pruned containers"), verbosity: :debug
         execute *KAMAL.prune.app_containers(retain: retain)
+        execute *KAMAL.prune.healthcheck_containers
       end
     end
   end

lib/kamal/cli/registry.rb

@@ -1,25 +1,17 @@
 class Kamal::Cli::Registry < Kamal::Cli::Base
-  desc "login", "Setup local registry or log in to remote registry locally and remotely"
+  desc "login", "Log in to registry locally and remotely"
   option :skip_local, aliases: "-L", type: :boolean, default: false, desc: "Skip local login"
   option :skip_remote, aliases: "-R", type: :boolean, default: false, desc: "Skip remote login"
-  def setup
-    if KAMAL.registry.local?
-      run_locally    { execute *KAMAL.registry.setup } unless options[:skip_local]
-    else
-      run_locally    { execute *KAMAL.registry.login } unless options[:skip_local]
-      on(KAMAL.hosts) { execute *KAMAL.registry.login } unless options[:skip_remote]
-    end
+  def login
+    run_locally    { execute *KAMAL.registry.login } unless options[:skip_local]
+    on(KAMAL.hosts) { execute *KAMAL.registry.login } unless options[:skip_remote]
   end
 
-  desc "remove", "Remove local registry or log out of remote registry locally and remotely"
+  desc "logout", "Log out of registry locally and remotely"
   option :skip_local, aliases: "-L", type: :boolean, default: false, desc: "Skip local login"
   option :skip_remote, aliases: "-R", type: :boolean, default: false, desc: "Skip remote login"
-  def remove
-    if KAMAL.registry.local?
-      run_locally    { execute *KAMAL.registry.remove, raise_on_non_zero_exit: false } unless options[:skip_local]
-    else
-      run_locally    { execute *KAMAL.registry.logout } unless options[:skip_local]
-      on(KAMAL.hosts) { execute *KAMAL.registry.logout } unless options[:skip_remote]
-    end
+  def logout
+    run_locally    { execute *KAMAL.registry.logout } unless options[:skip_local]
+    on(KAMAL.hosts) { execute *KAMAL.registry.logout } unless options[:skip_remote]
   end
 end

lib/kamal/cli/secrets.rb

@@ -5,20 +5,24 @@ class Kamal::Cli::Secrets < Kamal::Cli::Base
   option :from, type: :string, required: false, desc: "A vault or folder to fetch the secrets from"
   option :inline, type: :boolean, required: false, hidden: true
   def fetch(*secrets)
-    results = adapter(options[:adapter]).fetch(secrets, **options.slice(:account, :from).symbolize_keys)
-
-    return_or_puts JSON.dump(results).shellescape, inline: options[:inline]
+    handle_output(inline: options[:inline]) do
+      results = adapter(options[:adapter]).fetch(secrets, **options.slice(:account, :from).symbolize_keys)
+      JSON.dump(results).shellescape
+    end
   end
 
   desc "extract", "Extract a single secret from the results of a fetch call"
   option :inline, type: :boolean, required: false, hidden: true
   def extract(name, secrets)
-    parsed_secrets = JSON.parse(secrets)
-    value = parsed_secrets[name] || parsed_secrets.find { |k, v| k.end_with?("/#{name}") }&.last
+    handle_output(inline: options[:inline]) do
+      parsed_secrets = JSON.parse(secrets)
 
-    raise "Could not find secret #{name}" if value.nil?
+      value = parsed_secrets[name] || parsed_secrets.find { |k, v| k.end_with?("/#{name}") }&.last
 
-    return_or_puts value, inline: options[:inline]
+      raise "Could not find secret #{name}" if value.nil?
+
+      value
+    end
   end
 
   private
@@ -26,11 +30,18 @@ class Kamal::Cli::Secrets < Kamal::Cli::Base
       Kamal::Secrets::Adapters.lookup(adapter)
     end
 
-    def return_or_puts(value, inline: nil)
-      if inline
-        value
-      else
-        puts value
+    def handle_output(inline: nil)
+      yield.tap do |output|
+        puts output unless inline
       end
+    rescue => e
+      handle_error(e)
+    end
+
+    def handle_error(e)
+      $stderr.puts "  \e[31mERROR (#{e.class}): #{e.message}\e[0m"
+      $stderr.puts e.backtrace if ENV["VERBOSE"]
+
+      exit 1
     end
 end

lib/kamal/cli/server.rb

@@ -36,6 +36,8 @@ class Kamal::Cli::Server < Kamal::Cli::Base
             missing << host
           end
         end
+
+        execute(*KAMAL.server.ensure_run_directory)
       end
 
       if missing.any?

lib/kamal/cli/templates/deploy.yml

@@ -57,6 +57,17 @@ builder:
 #     directories:
 #       - data:/data
 
+# Configure custom arguments for Traefik. Be sure to reboot traefik when you modify it.
+# traefik:
+#   args:
+#     accesslog: true
+#     accesslog.format: json
+
+# Configure a custom healthcheck (default is /up on port 3000)
+# healthcheck:
+#   path: /healthz
+#   port: 4000
+
 # 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.

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

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

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/traefik.rb

@@ -0,0 +1,141 @@
+class Kamal::Cli::Traefik < Kamal::Cli::Base
+  desc "boot", "Boot Traefik on servers"
+  def boot
+    raise_if_kamal_proxy_enabled!
+    with_lock do
+      on(KAMAL.traefik_hosts) do
+        execute *KAMAL.registry.login
+        execute *KAMAL.traefik.ensure_env_directory
+        upload! KAMAL.traefik.secrets_io, KAMAL.traefik.secrets_path, mode: "0600"
+        execute *KAMAL.traefik.start_or_run
+      end
+    end
+  end
+
+  desc "reboot", "Reboot Traefik on servers (stop container, remove container, start new container)"
+  option :rolling, type: :boolean, default: false, desc: "Reboot traefik on hosts in sequence, rather than in parallel"
+  option :confirmed, aliases: "-y", type: :boolean, default: false, desc: "Proceed without confirmation question"
+  def reboot
+    raise_if_kamal_proxy_enabled!
+    confirming "This will cause a brief outage on each host. Are you sure?" do
+      with_lock 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, 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
+  end
+
+  desc "start", "Start existing Traefik container on servers"
+  def start
+    raise_if_kamal_proxy_enabled!
+    with_lock do
+      on(KAMAL.traefik_hosts) do
+        execute *KAMAL.auditor.record("Started traefik"), verbosity: :debug
+        execute *KAMAL.traefik.start
+      end
+    end
+  end
+
+  desc "stop", "Stop existing Traefik container on servers"
+  def stop
+    raise_if_kamal_proxy_enabled!
+    with_lock do
+      on(KAMAL.traefik_hosts) do
+        execute *KAMAL.auditor.record("Stopped traefik"), verbosity: :debug
+        execute *KAMAL.traefik.stop, raise_on_non_zero_exit: false
+      end
+    end
+  end
+
+  desc "restart", "Restart existing Traefik container on servers"
+  def restart
+    raise_if_kamal_proxy_enabled!
+    with_lock do
+      stop
+      start
+    end
+  end
+
+  desc "details", "Show details about Traefik container from servers"
+  def details
+    raise_if_kamal_proxy_enabled!
+    on(KAMAL.traefik_hosts) { |host| puts_by_host host, capture_with_info(*KAMAL.traefik.info), type: "Traefik" }
+  end
+
+  desc "logs", "Show log lines from Traefik on servers"
+  option :since, aliases: "-s", desc: "Show logs since timestamp (e.g. 2013-01-02T13:23:37Z) or relative (e.g. 42m for 42 minutes)"
+  option :lines, type: :numeric, aliases: "-n", desc: "Number of log lines to pull from each server"
+  option :grep, aliases: "-g", desc: "Show lines with grep match only (use this to fetch specific requests by id)"
+  option :grep_options, aliases: "-o", desc: "Additional options supplied to grep"
+  option :follow, aliases: "-f", desc: "Follow logs on primary server (or specific host set by --hosts)"
+  def logs
+    raise_if_kamal_proxy_enabled!
+    grep = options[:grep]
+    grep_options = options[:grep_options]
+
+    if options[:follow]
+      run_locally do
+        info "Following logs on #{KAMAL.primary_host}..."
+        info KAMAL.traefik.follow_logs(host: KAMAL.primary_host, grep: grep, grep_options: grep_options)
+        exec KAMAL.traefik.follow_logs(host: KAMAL.primary_host, grep: grep, grep_options: grep_options)
+      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.traefik_hosts) do |host|
+        puts_by_host host, capture(*KAMAL.traefik.logs(since: since, lines: lines, grep: grep, grep_options: grep_options)), type: "Traefik"
+      end
+    end
+  end
+
+  desc "remove", "Remove Traefik container and image from servers"
+  def remove
+    raise_if_kamal_proxy_enabled!
+    with_lock do
+      stop
+      remove_container
+      remove_image
+    end
+  end
+
+  desc "remove_container", "Remove Traefik container from servers", hide: true
+  def remove_container
+    raise_if_kamal_proxy_enabled!
+    with_lock do
+      on(KAMAL.traefik_hosts) do
+        execute *KAMAL.auditor.record("Removed traefik container"), verbosity: :debug
+        execute *KAMAL.traefik.remove_container
+      end
+    end
+  end
+
+  desc "remove_image", "Remove Traefik image from servers", hide: true
+  def remove_image
+    raise_if_kamal_proxy_enabled!
+    with_lock do
+      on(KAMAL.traefik_hosts) do
+        execute *KAMAL.auditor.record("Removed traefik image"), verbosity: :debug
+        execute *KAMAL.traefik.remove_image
+      end
+    end
+  end
+
+  private
+    def raise_if_kamal_proxy_enabled!
+      if KAMAL.config.proxy.enabled?
+        raise "kamal traefik commands are disabled when experimental proxy support is enabled. Use `kamal proxy` commands instead."
+      end
+    end
+end

lib/kamal/cli/tunnel/remote_ports.rb

@@ -1,66 +0,0 @@
-Signal.trap "SIGPROF" do
-  Thread.list.each do |thread|
-    puts thread.name
-    puts thread.backtrace.map { |bt| "  #{bt}" }
-    puts
-  end
-end
-
-require "concurrent/map"
-
-class Kamal::Cli::Tunnel::RemotePorts
-  attr_reader :hosts, :port
-
-  def initialize(hosts, port)
-    @hosts = hosts
-    @port = port
-    @open = false
-  end
-
-  def open
-    @open = true
-    @opened = Concurrent::Map.new
-
-    @threads = hosts.map do |host|
-      Thread.new do
-        Net::SSH.start(host, KAMAL.config.ssh.user) do |ssh|
-          forwarding = nil
-          ssh.forward.remote(port, "localhost", port, "localhost") do |actual_remote_port|
-            forwarding = !!actual_remote_port
-            :no_exception # will yield the exception on my own thread
-          end
-          ssh.loop { forwarding.nil? }
-          if forwarding
-            @opened[host] = true
-            ssh.loop(0.1) { @open }
-
-            ssh.forward.cancel_remote(port, "localhost")
-            ssh.loop(0.1) { ssh.forward.active_remotes.include?([ port, "localhost" ]) }
-          else
-            @opened[host] = false
-          end
-        end
-      rescue => e
-        @opened[host] = false
-
-        puts e.message
-        puts e.backtrace
-      end
-    end
-
-    loop do
-      break if @opened.size == hosts.size
-      sleep 0.1
-    end
-
-    raise "Could not open tunnels on #{opened.reject { |k, v| v }.join(", ")}" unless @opened.values.all?
-  end
-
-  def close
-    p "Closing"
-    @open = false
-    p "Joining"
-    @threads.each(&:join)
-    p "Joined"
-  end
-end

lib/kamal/commander.rb

@@ -4,7 +4,7 @@ require "active_support/core_ext/object/blank"
 
 class Kamal::Commander
   attr_accessor :verbosity, :holding_lock, :connected
-  delegate :hosts, :roles, :primary_host, :primary_role, :roles_on, :proxy_hosts, :accessory_hosts, to: :specifics
+  delegate :hosts, :roles, :primary_host, :primary_role, :roles_on, :traefik_hosts, :proxy_hosts, :proxy_host?, :accessory_hosts, to: :specifics
 
   def initialize
     self.verbosity = :info
@@ -65,13 +65,6 @@ class Kamal::Commander
     end
   end
 
-  def with_specific_hosts(hosts)
-    original_hosts, self.specific_hosts = specific_hosts, hosts
-    yield
-  ensure
-    self.specific_hosts = original_hosts
-  end
-
   def accessory_names
     config.accessories&.collect(&:name) || []
   end
@@ -101,6 +94,10 @@ class Kamal::Commander
     @docker ||= Kamal::Commands::Docker.new(config)
   end
 
+  def healthcheck
+    @healthcheck ||= Kamal::Commands::Healthcheck.new(config)
+  end
+
   def hook
     @hook ||= Kamal::Commands::Hook.new(config)
   end
@@ -125,11 +122,19 @@ class Kamal::Commander
     @server ||= Kamal::Commands::Server.new(config)
   end
 
+  def traefik
+    @traefik ||= Kamal::Commands::Traefik.new(config)
+  end
+
   def alias(name)
     config.aliases[name]
   end
 
 
+  def traefik_or_proxy(host)
+    proxy_host?(host) ? proxy : traefik
+  end
+
   def with_verbosity(level)
     old_level = self.verbosity
 

lib/kamal/commander/specifics.rb

@@ -18,12 +18,21 @@ class Kamal::Commander::Specifics
     roles.select { |role| role.hosts.include?(host.to_s) }
   end
 
+  def traefik_hosts
+    config.traefik_hosts & specified_hosts
+  end
+
   def proxy_hosts
-    config.proxy_hosts & specified_hosts
+    config.proxy_hosts
+  end
+
+  def proxy_host?(host)
+    host = host.hostname if host.is_a?(SSHKit::Host)
+    proxy_hosts.include?(host)
   end
 
   def accessory_hosts
-    config.accessories.flat_map(&:hosts) & specified_hosts
+    specific_hosts || config.accessories.flat_map(&:hosts)
   end
 
   private

lib/kamal/commands/accessory.rb

@@ -39,16 +39,16 @@ class Kamal::Commands::Accessory < Kamal::Commands::Base
   end
 
 
-  def logs(timestamps: true, since: nil, lines: nil, grep: nil, grep_options: nil)
+  def logs(since: nil, lines: nil, grep: nil, grep_options: nil)
     pipe \
-      docker(:logs, service_name, (" --since #{since}" if since), (" --tail #{lines}" if lines), ("--timestamps" if timestamps), "2>&1"),
+      docker(:logs, service_name, (" --since #{since}" if since), (" --tail #{lines}" if lines), "--timestamps", "2>&1"),
       ("grep '#{grep}'#{" #{grep_options}" if grep_options}" if grep)
   end
 
-  def follow_logs(timestamps: true, grep: nil, grep_options: nil)
+  def follow_logs(grep: nil, grep_options: nil)
     run_over_ssh \
       pipe \
-        docker(:logs, service_name, ("--timestamps" if timestamps), "--tail", "10", "--follow", "2>&1"),
+        docker(:logs, service_name, "--timestamps", "--tail", "10", "--follow", "2>&1"),
         (%(grep "#{grep}"#{" #{grep_options}" if grep_options}) if grep)
   end
 

lib/kamal/commands/app.rb

@@ -1,5 +1,5 @@
 class Kamal::Commands::App < Kamal::Commands::Base
-  include Assets, Containers, Execution, Images, Logging, Proxy
+  include Assets, Containers, Cord, Execution, Images, Logging
 
   ACTIVE_DOCKER_STATUSES = [ :running, :restarting ]
 
@@ -14,6 +14,25 @@ class Kamal::Commands::App < Kamal::Commands::Base
   end
 
   def run(hostname: nil)
+    docker :run,
+      "--detach",
+      "--restart unless-stopped",
+      "--name", container_name,
+      *([ "--hostname", hostname ] if hostname),
+      "-e", "KAMAL_CONTAINER_NAME=\"#{container_name}\"",
+      "-e", "KAMAL_VERSION=\"#{config.version}\"",
+      *role.env_args(host),
+      *role.health_check_args,
+      *role.logging_args,
+      *config.volume_args,
+      *role.asset_volume_args,
+      *role.label_args,
+      *role.option_args,
+      config.absolute_image,
+      role.cmd
+  end
+
+  def run_for_proxy(hostname: nil)
     docker :run,
       "--detach",
       "--restart unless-stopped",
@@ -26,7 +45,7 @@ class Kamal::Commands::App < Kamal::Commands::Base
       *role.logging_args,
       *config.volume_args,
       *role.asset_volume_args,
-      *role.label_args,
+      *role.label_args_for_proxy,
       *role.option_args,
       config.absolute_image,
       role.cmd
@@ -43,7 +62,7 @@ class Kamal::Commands::App < Kamal::Commands::Base
   def stop(version: nil)
     pipe \
       version ? container_id_for_version(version) : current_running_container_id,
-      xargs(docker(:stop, *role.stop_args))
+      xargs(config.stop_wait_time ? docker(:stop, "-t", config.stop_wait_time) : docker(:stop))
   end
 
   def info

lib/kamal/commands/app/assets.rb

@@ -3,18 +3,18 @@ module Kamal::Commands::App::Assets
     asset_container = "#{role.container_prefix}-assets"
 
     combine \
-      make_directory(role.asset_extracted_directory),
+      make_directory(role.asset_extracted_path),
       [ *docker(:stop, "-t 1", asset_container, "2> /dev/null"), "|| true" ],
-      docker(:run, "--name", asset_container, "--detach", "--rm", "--entrypoint", "sleep", config.absolute_image, "1000000"),
-      docker(:cp, "-L", "#{asset_container}:#{role.asset_path}/.", role.asset_extracted_directory),
+      docker(:run, "--name", asset_container, "--detach", "--rm", config.absolute_image, "sleep 1000000"),
+      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.asset_extracted_directory(config.version), role.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.asset_extracted_directory(old_version), role.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.asset_extracted_directory),
-      find_and_remove_older_siblings(role.asset_volume_directory)
+      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.name}-*'",
+        "-name", "'#{role.container_prefix}-*'",
         "!", "-name", Pathname.new(path).basename.to_s,
         "-exec rm -rf \"{}\" +"
       ]

lib/kamal/commands/app/cord.rb

@@ -0,0 +1,22 @@
+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.cord_volume.container_path}\" {print $1}'" ]
+  end
+
+  def tie_cord(cord)
+    create_empty_file(cord)
+  end
+
+  def cut_cord(cord)
+    remove_directory(cord)
+  end
+
+  private
+    def create_empty_file(file)
+      chain \
+        make_directory_for(file),
+        [ :touch, file ]
+    end
+end

lib/kamal/commands/app/execution.rb

@@ -11,7 +11,6 @@ module Kamal::Commands::App::Execution
     docker :run,
       ("-it" if interactive),
       "--rm",
-      "--network", "kamal",
       *role&.env_args(host),
       *argumentize("--env", env),
       *config.volume_args,

lib/kamal/commands/app/logging.rb

@@ -1,16 +1,16 @@
 module Kamal::Commands::App::Logging
-  def logs(version: nil, timestamps: true, since: nil, lines: nil, grep: nil, grep_options: nil)
+  def logs(version: nil, since: nil, lines: nil, grep: nil, grep_options: nil)
     pipe \
       version ? container_id_for_version(version) : current_running_container_id,
-      "xargs docker logs#{" --timestamps" if timestamps}#{" --since #{since}" if since}#{" --tail #{lines}" if lines} 2>&1",
+      "xargs docker logs#{" --since #{since}" if since}#{" --tail #{lines}" if lines} 2>&1",
       ("grep '#{grep}'#{" #{grep_options}" if grep_options}" if grep)
   end
 
-  def follow_logs(host:, timestamps: true, lines: nil, grep: nil, grep_options: nil)
+  def follow_logs(host:, lines: nil, grep: nil, grep_options: nil)
     run_over_ssh \
       pipe(
         current_running_container_id,
-        "xargs docker logs#{" --timestamps" if timestamps}#{" --tail #{lines}" if lines} --follow 2>&1",
+        "xargs docker logs --timestamps#{" --tail #{lines}" if lines} --follow 2>&1",
         (%(grep "#{grep}"#{" #{grep_options}" if grep_options}) if grep)
       ),
       host: host

lib/kamal/commands/app/proxy.rb

@@ -1,16 +0,0 @@
-module Kamal::Commands::App::Proxy
-  delegate :proxy_container_name, to: :config
-
-  def deploy(target:)
-    proxy_exec :deploy, role.container_prefix, *role.proxy.deploy_command_args(target: target)
-  end
-
-  def remove(target:)
-    proxy_exec :remove, role.container_prefix, *role.proxy.remove_command_args(target: target)
-  end
-
-  private
-    def proxy_exec(*command)
-      docker :exec, proxy_container_name, "kamal-proxy", *command
-    end
-end

lib/kamal/commands/hook.rb

@@ -1,12 +1,9 @@
 class Kamal::Commands::Hook < Kamal::Commands::Base
-  def run(hook)
-    [ hook_file(hook) ]
-  end
+  def run(hook, secrets: false, **details)
+    env = tags(**details).env
+    env.merge!(config.secrets.to_h) if secrets
 
-  def env(secrets: false, **details)
-    tags(**details).env.tap do |env|
-      env.merge!(config.secrets.to_h) if secrets
-    end
+    [ hook_file(hook), env: env ]
   end
 
   def hook_exists?(hook)

lib/kamal/commands/lock.rb

@@ -44,10 +44,14 @@ class Kamal::Commands::Lock < Kamal::Commands::Base
         "/dev/null"
     end
 
-    def lock_dir
-      dir_name = [ "lock", config.service, config.destination ].compact.join("-")
+    def locks_dir
+      File.join(config.run_directory, "locks")
+    end
 
-      File.join(config.run_directory, dir_name)
+    def lock_dir
+      dir_name = [ config.service, config.destination ].compact.join("-")
+
+      File.join(locks_dir, dir_name)
     end
 
     def lock_details_file

lib/kamal/commands/proxy.rb

@@ -1,5 +1,13 @@
 class Kamal::Commands::Proxy < Kamal::Commands::Base
   delegate :argumentize, :optionize, to: Kamal::Utils
+  delegate :container_name, :app_port, to: :proxy_config
+
+  attr_reader :proxy_config
+
+  def initialize(config)
+    super
+    @proxy_config = config.proxy
+  end
 
   def run
     docker :run,
@@ -7,10 +15,11 @@ class Kamal::Commands::Proxy < Kamal::Commands::Base
       "--network", "kamal",
       "--detach",
       "--restart", "unless-stopped",
-      *config.proxy_publish_args,
-      "--volume", "kamal-proxy-config:/home/kamal-proxy/.config/kamal-proxy",
+      *proxy_config.publish_args,
+      "--volume", "/var/run/docker.sock:/var/run/docker.sock",
+      "--volume", "#{proxy_config.config_directory_as_docker_volume}:/root/.config/kamal-proxy",
       *config.logging_args,
-      config.proxy_image
+      proxy_config.image
   end
 
   def start
@@ -25,25 +34,27 @@ class Kamal::Commands::Proxy < Kamal::Commands::Base
     combine start, run, by: "||"
   end
 
+  def deploy(service, target:)
+    docker :exec, container_name, "kamal-proxy", :deploy, service, *optionize({ target: "#{target}:#{app_port}" }), *proxy_config.deploy_command_args
+  end
+
+  def remove(service, target:)
+    docker :exec, container_name, "kamal-proxy", :remove, service, *optionize({ target: "#{target}:#{app_port}" })
+  end
+
   def info
     docker :ps, "--filter", "name=^#{container_name}$"
   end
 
-  def version
+  def logs(since: nil, lines: nil, grep: nil, grep_options: nil)
     pipe \
-      docker(:inspect, container_name, "--format '{{.Config.Image}}'"),
-      [ :cut, "-d:", "-f2" ]
-  end
-
-  def logs(timestamps: true, since: nil, lines: nil, grep: nil, grep_options: nil)
-    pipe \
-      docker(:logs, container_name, ("--since #{since}" if since), ("--tail #{lines}" if lines), ("--timestamps" if timestamps), "2>&1"),
+      docker(:logs, container_name, (" --since #{since}" if since), (" --tail #{lines}" if lines), "--timestamps", "2>&1"),
       ("grep '#{grep}'#{" #{grep_options}" if grep_options}" if grep)
   end
 
-  def follow_logs(host:, timestamps: true, grep: nil, grep_options: nil)
+  def follow_logs(host:, grep: nil, grep_options: nil)
     run_over_ssh pipe(
-      docker(:logs, container_name, ("--timestamps" if timestamps), "--tail", "10", "--follow", "2>&1"),
+      docker(:logs, container_name, "--timestamps", "--tail", "10", "--follow", "2>&1"),
       (%(grep "#{grep}"#{" #{grep_options}" if grep_options}) if grep)
     ).join(" "), host: host
   end
@@ -55,18 +66,4 @@ class Kamal::Commands::Proxy < Kamal::Commands::Base
   def remove_image
     docker :image, :prune, "--all", "--force", "--filter", "label=org.opencontainers.image.title=kamal-proxy"
   end
-
-  def cleanup_traefik
-    chain \
-      docker(:container, :stop, "traefik"),
-      combine(
-        docker(:container, :prune, "--force", "--filter", "label=org.opencontainers.image.title=Traefik"),
-        docker(:image, :prune, "--all", "--force", "--filter", "label=org.opencontainers.image.title=Traefik")
-      )
-  end
-
-  private
-    def container_name
-      config.proxy_container_name
-    end
 end

lib/kamal/commands/prune.rb

@@ -20,6 +20,10 @@ class Kamal::Commands::Prune < Kamal::Commands::Base
       "while read container_id; do docker rm $container_id; done"
   end
 
+  def healthcheck_containers
+    docker :container, :prune, "--force", *healthcheck_service_filter
+  end
+
   private
     def stopped_containers_filters
       [ "created", "exited", "dead" ].flat_map { |status| [ "--filter", "status=#{status}" ] }
@@ -35,4 +39,8 @@ class Kamal::Commands::Prune < Kamal::Commands::Base
     def service_filter
       [ "--filter", "label=service=#{config.service}" ]
     end
+
+    def healthcheck_service_filter
+      [ "--filter", "label=service=#{config.healthcheck_service}" ]
+    end
 end

lib/kamal/commands/registry.rb

@@ -1,6 +1,5 @@
 class Kamal::Commands::Registry < Kamal::Commands::Base
   delegate :registry, to: :config
-  delegate :local?, :local_port, to: :registry
 
   def login
     docker :login,
@@ -12,26 +11,4 @@ class Kamal::Commands::Registry < Kamal::Commands::Base
   def logout
     docker :logout, registry.server
   end
-
-  def setup
-    combine \
-      docker(:start, "kamal-docker-registry"),
-      docker(:run, "--detach", "-p", "127.0.0.1:#{local_port}:5000", "--name", "kamal-docker-registry", "registry:2"),
-      by: "||"
-  end
-
-  def remove
-    combine \
-      docker(:stop, "kamal-docker-registry"),
-      docker(:rm, "kamal-docker-registry"),
-      by: "&&"
-  end
-
-  def logout
-    docker :logout, registry.server
-  end
-
-  def tunnel(host)
-    run_over_ssh "-R", "#{local_port}:localhost:#{local_port}", host: host
-  end
 end

lib/kamal/commands/server.rb

@@ -1,15 +1,5 @@
 class Kamal::Commands::Server < Kamal::Commands::Base
   def ensure_run_directory
-    make_directory config.run_directory
-  end
-
-  def remove_app_directory
-    remove_directory config.app_directory
-  end
-
-  def app_directory_count
-    pipe \
-      [ :ls, config.apps_directory ],
-      [ :wc, "-l" ]
+    [ :mkdir, "-p", config.run_directory ]
   end
 end

lib/kamal/commands/traefik.rb

@@ -0,0 +1,77 @@
+class Kamal::Commands::Traefik < Kamal::Commands::Base
+  delegate :argumentize, :optionize, to: Kamal::Utils
+  delegate :port, :publish?, :labels, :env, :image, :options, :args, :env_args, :secrets_io, :env_directory, :secrets_path, to: :"config.traefik"
+
+  def run
+    docker :run, "--name traefik",
+      "--detach",
+      "--restart", "unless-stopped",
+      *publish_args,
+      "--volume", "/var/run/docker.sock:/var/run/docker.sock",
+      *env_args,
+      *config.logging_args,
+      *label_args,
+      *docker_options_args,
+      image,
+      "--providers.docker",
+      *cmd_option_args
+  end
+
+  def start
+    docker :container, :start, "traefik"
+  end
+
+  def stop
+    docker :container, :stop, "traefik"
+  end
+
+  def start_or_run
+    any start, run
+  end
+
+  def info
+    docker :ps, "--filter", "name=^traefik$"
+  end
+
+  def logs(since: nil, lines: nil, grep: nil, grep_options: nil)
+    pipe \
+      docker(:logs, "traefik", (" --since #{since}" if since), (" --tail #{lines}" if lines), "--timestamps", "2>&1"),
+      ("grep '#{grep}'#{" #{grep_options}" if grep_options}" if grep)
+  end
+
+  def follow_logs(host:, grep: nil, grep_options: nil)
+    run_over_ssh pipe(
+      docker(:logs, "traefik", "--timestamps", "--tail", "10", "--follow", "2>&1"),
+      (%(grep "#{grep}"#{" #{grep_options}" if grep_options}) if grep)
+    ).join(" "), host: host
+  end
+
+  def remove_container
+    docker :container, :prune, "--force", "--filter", "label=org.opencontainers.image.title=Traefik"
+  end
+
+  def remove_image
+    docker :image, :prune, "--all", "--force", "--filter", "label=org.opencontainers.image.title=Traefik"
+  end
+
+  def ensure_env_directory
+    make_directory env_directory
+  end
+
+  private
+    def publish_args
+      argumentize "--publish", port if publish?
+    end
+
+    def label_args
+      argumentize "--label", labels
+    end
+
+    def docker_options_args
+      optionize(options)
+    end
+
+    def cmd_option_args
+      optionize args, with: "="
+    end
+end

lib/kamal/configuration.rb

@@ -6,18 +6,14 @@ require "erb"
 require "net/ssh/proxy/jump"
 
 class Kamal::Configuration
-  delegate :service, :image, :labels, :hooks_path, to: :raw_config, allow_nil: true
+  delegate :service, :image, :labels, :stop_wait_time, :hooks_path, to: :raw_config, allow_nil: true
   delegate :argumentize, :optionize, to: Kamal::Utils
 
-  attr_reader :destination, :raw_config, :secrets
-  attr_reader :accessories, :aliases, :boot, :builder, :env, :logging, :proxy, :servers, :ssh, :sshkit, :registry
+  attr_reader :destination, :raw_config
+  attr_reader :accessories, :aliases, :boot, :builder, :env, :healthcheck, :logging, :proxy, :traefik, :servers, :ssh, :sshkit, :registry
 
   include Validation
 
-  PROXY_MINIMUM_VERSION = "v0.4.0"
-  PROXY_HTTP_PORT = 80
-  PROXY_HTTPS_PORT = 443
-
   class << self
     def create_from(config_file:, destination: nil, version: nil)
       raw_config = load_config_files(config_file, *destination_config_file(config_file, destination))
@@ -52,8 +48,6 @@ class Kamal::Configuration
 
     validate! raw_config, example: validation_yml.symbolize_keys, context: "", with: Kamal::Configuration::Validator::Configuration
 
-    @secrets = Kamal::Secrets.new(destination: destination)
-
     # Eager load config to validate it, these are first as they have dependencies later on
     @servers = Servers.new(config: self)
     @registry = Registry.new(config: self)
@@ -64,8 +58,10 @@ class Kamal::Configuration
     @builder = Builder.new(config: self)
     @env = Env.new(config: @raw_config.env || {}, secrets: secrets)
 
+    @healthcheck = Healthcheck.new(healthcheck_config: @raw_config.healthcheck)
     @logging = Logging.new(logging_config: @raw_config.logging)
-    @proxy = Proxy.new(config: self, proxy_config: @raw_config.proxy || {})
+    @proxy = Proxy.new(config: self)
+    @traefik = Traefik.new(config: self)
     @ssh = Ssh.new(config: self)
     @sshkit = Sshkit.new(config: self)
 
@@ -74,9 +70,6 @@ class Kamal::Configuration
     ensure_valid_kamal_version
     ensure_retain_containers_valid
     ensure_valid_service_name
-    ensure_no_traefik_reboot_hooks
-    ensure_one_host_for_ssl_roles
-    ensure_unique_hosts_for_ssl_roles
   end
 
 
@@ -137,16 +130,20 @@ class Kamal::Configuration
     raw_config.allow_empty_roles
   end
 
-  def proxy_roles
-    roles.select(&:running_proxy?)
+  def traefik_roles
+    roles.select(&:running_traefik?)
   end
 
-  def proxy_role_names
-    proxy_roles.flat_map(&:name)
+  def traefik_role_names
+    traefik_roles.flat_map(&:name)
+  end
+
+  def traefik_hosts
+    traefik_roles.flat_map(&:hosts).uniq
   end
 
   def proxy_hosts
-    proxy_roles.flat_map(&:hosts).uniq
+    proxy.hosts
   end
 
   def repository
@@ -191,16 +188,16 @@ class Kamal::Configuration
   end
 
 
+  def healthcheck_service
+    [ "healthcheck", service, destination ].compact.join("-")
+  end
+
   def readiness_delay
     raw_config.readiness_delay || 7
   end
 
-  def deploy_timeout
-    raw_config.deploy_timeout || 30
-  end
-
-  def drain_timeout
-    raw_config.drain_timeout || 30
+  def run_id
+    @run_id ||= SecureRandom.hex(16)
   end
 
 
@@ -208,23 +205,10 @@ class Kamal::Configuration
     ".kamal"
   end
 
-  def apps_directory
-    File.join run_directory, "apps"
+  def run_directory_as_docker_volume
+    File.join "$(pwd)", run_directory
   end
 
-  def app_directory
-    File.join apps_directory, [ service, destination ].compact.join("-")
-  end
-
-  def env_directory
-    File.join app_directory, "env"
-  end
-
-  def assets_directory
-    File.join app_directory, "assets"
-  end
-
-
   def hooks_path
     raw_config.hooks_path || ".kamal/hooks"
   end
@@ -234,6 +218,10 @@ class Kamal::Configuration
   end
 
 
+  def env_directory
+    File.join(run_directory, "env")
+  end
+
   def env_tags
     @env_tags ||= if (tags = raw_config.env["tags"])
       tags.collect { |name, config| Env::Tag.new(name, config: config, secrets: secrets) }
@@ -246,18 +234,6 @@ class Kamal::Configuration
     env_tags.detect { |t| t.name == name.to_s }
   end
 
-  def proxy_publish_args
-    argumentize "--publish", [ "#{PROXY_HTTP_PORT}:#{PROXY_HTTP_PORT}", "#{PROXY_HTTPS_PORT}:#{PROXY_HTTPS_PORT}" ]
-  end
-
-  def proxy_image
-    "basecamp/kamal-proxy:#{PROXY_MINIMUM_VERSION}"
-  end
-
-  def proxy_container_name
-    "kamal-proxy"
-  end
-
 
   def to_h
     {
@@ -273,10 +249,15 @@ class Kamal::Configuration
       sshkit: sshkit.to_h,
       builder: builder.to_h,
       accessories: raw_config.accessories,
-      logging: logging_args
+      logging: logging_args,
+      healthcheck: healthcheck.to_h
     }.compact
   end
 
+  def secrets
+    @secrets ||= Kamal::Secrets.new(destination: destination)
+  end
+
   private
     # Will raise ArgumentError if any required config keys are missing
     def ensure_destination_if_required
@@ -331,30 +312,6 @@ class Kamal::Configuration
       true
     end
 
-    def ensure_no_traefik_reboot_hooks
-      hooks = %w[ pre-traefik-reboot post-traefik-reboot ].select { |hook_file| File.exist?(File.join(hooks_path, hook_file)) }
-
-      if hooks.any?
-        raise Kamal::ConfigurationError, "Found #{hooks.join(", ")}, these should be renamed to (pre|post)-proxy-reboot"
-      end
-
-      true
-    end
-
-    def ensure_one_host_for_ssl_roles
-      roles.each(&:ensure_one_host_for_ssl)
-
-      true
-    end
-
-    def ensure_unique_hosts_for_ssl_roles
-      hosts = roles.select(&:ssl?).map { |role| role.proxy.host }
-      duplicates = hosts.tally.filter_map { |host, count| host if count > 1 }
-
-      raise Kamal::ConfigurationError, "Different roles can't share the same host for SSL: #{duplicates.join(", ")}" if duplicates.any?
-
-      true
-    end
 
     def role_names
       raw_config.servers.is_a?(Array) ? [ "web" ] : raw_config.servers.keys.sort

lib/kamal/configuration/accessory.rb

@@ -63,7 +63,7 @@ class Kamal::Configuration::Accessory
   end
 
   def secrets_path
-    File.join(config.env_directory, "accessories", "#{name}.env")
+    File.join(config.env_directory, "accessories", "#{service_name}.env")
   end
 
   def files

lib/kamal/configuration/docs/builder.yml

@@ -2,31 +2,35 @@
 #
 # The builder configuration controls how the application is built with `docker build`
 #
+# If no configuration is specified, Kamal will:
+# 1. Create a buildx context called `kamal-local-docker-container`, using the docker-container driver
+# 2. Use `docker build` to build a multiarch image for linux/amd64,linux/arm64 with that context
+#
 # See https://kamal-deploy.org/docs/configuration/builder-examples/ for more information
 
 # Builder options
 #
 # Options go under the builder key in the root configuration.
 builder:
+  # Driver
+  #
+  # The build driver to use, defaults to `docker-container`
+  driver: docker
 
   # Arch
   #
-  # The architectures to build for - you can set an array or just a single value.
-  #
-  # Allowed values are `amd64` and `arm64`
+  # The architectures to build for, defaults to `[ amd64, arm64 ]`
+  # Unless you are using the docker driver, when it defaults to the local architecture
+  # You can set an array or just a single value
   arch:
     - amd64
 
-  # Remote
+  # Remote configuration
   #
-  # The connection string for a remote builder. If supplied Kamal will use this
-  # for builds that do not match the local architecture of the deployment host.
+  # If you have a remote builder, you can configure it here
   remote: ssh://docker@docker-builder
 
-  # Local
-  #
-  # If set to false, Kamal will always use the remote builder even when building
-  # the local architecture.
+  # Whether to allow local builds
   #
   # Defaults to true
   local: true
@@ -74,7 +78,7 @@ builder:
 
   # Build secrets
   #
-  # Values are read from .kamal/secrets.
+  # Values are read from the environment.
   #
   secrets:
     - SECRET1
@@ -99,8 +103,3 @@ builder:
   #
   # SSH agent socket or keys to expose to the build
   ssh: default=$SSH_AUTH_SOCK
-
-  # Driver
-  #
-  # The build driver to use, defaults to `docker-container`
-  driver: docker

lib/kamal/configuration/docs/configuration.yml

@@ -36,8 +36,6 @@ image: my-image
 labels:
   my-label: my-value
 
-# Volumes
-#
 # Additional volumes to mount into the container
 volumes:
   - /path/on/host:/path/in/container:ro
@@ -60,7 +58,7 @@ servers:
 env:
   ...
 
-# Asset Path
+# Asset Bridging
 #
 # Used for asset bridging across deployments, default to `nil`
 #
@@ -72,12 +70,10 @@ env:
 # volume containing both sets of files.
 # This requires that file names change when the contents change
 # (e.g. by including a hash of the contents in the name).
-#
+
 # To configure this, set the path to the assets:
 asset_path: /path/to/assets
 
-# Hooks path
-#
 # Path to hooks, defaults to `.kamal/hooks`
 # See https://kamal-deploy.org/docs/hooks for more information
 hooks_path: /user_home/kamal/hooks
@@ -87,7 +83,7 @@ hooks_path: /user_home/kamal/hooks
 # Whether deployments require a destination to be specified, defaults to `false`
 require_destination: true
 
-# Primary role
+# The primary role
 #
 # This defaults to `web`, but if you have no web role, you can change this
 primary_role: workers
@@ -97,6 +93,11 @@ primary_role: workers
 # Whether roles with no servers are allowed. Defaults to `false`.
 allow_empty_roles: false
 
+# Stop wait time
+#
+# How long we wait for a container to stop before killing it, defaults to 30 seconds
+stop_wait_time: 60
+
 # Retain containers
 #
 # How many old containers and images we retain, defaults to 5
@@ -110,20 +111,9 @@ minimum_version: 1.3.0
 # Readiness delay
 #
 # Seconds to wait for a container to boot after is running, default 7
-#
-# This only applies to containers that do not run a proxy or specify a healthcheck
+# This only applies to containers that do not specify a healthcheck
 readiness_delay: 4
 
-# Deploy timeout
-#
-# How long to wait for a container to become ready, default 30
-deploy_timeout: 10
-
-# Drain timeout
-#
-# How long to wait for a containers to drain, default 30
-drain_timeout: 10
-
 # Run directory
 #
 # Directory to store kamal runtime files in on the host, default `.kamal`
@@ -147,9 +137,15 @@ builder:
 accessories:
   ...
 
+# Traefik
+#
+# The Traefik proxy is used for zero-downtime deployments, see kamal docs traefik
+traefik:
+  ...
+
 # Proxy
 #
-# Configuration for kamal-proxy, see kamal docs proxy
+# **Experimental** Configuration for kamal-proxy the replacement for Traefik, see kamal docs proxy
 proxy:
   ...
 
@@ -165,6 +161,12 @@ sshkit:
 boot:
   ...
 
+# Healthcheck
+#
+# Configuring healthcheck commands, intervals and timeouts, see kamal docs healthcheck
+healthcheck:
+  ...
+
 # Logging
 #
 # Docker logging configuration, see kamal docs logging

lib/kamal/configuration/docs/env.yml

@@ -1,7 +1,7 @@
 # Environment variables
 #
 # Environment variables can be set directly in the Kamal configuration or
-# read from .kamal/secrets.
+# loaded from a .env file, for secrets that should not be checked into Git.
 
 # Reading environment variables from the configuration
 #
@@ -12,33 +12,19 @@ env:
   DATABASE_HOST: mysql-db1
   DATABASE_PORT: 3306
 
-# Secrets
+# Using .env file to load required environment variables
 #
-# Kamal uses dotenv to automatically load environment variables set in the `.kamal/secrets` file.
-#
-# If you are using destinations, secrets will instead be read from `.kamal/secrets-<DESTINATION>` if
-# it exists.
-#
-# Common secrets across all destinations can be set in `.kamal/secrets-common`.
-#
-# This file can be used to set variables like `KAMAL_REGISTRY_PASSWORD` or database passwords.
-# You can use variable or command substitution in the secrets file.
+# Kamal uses dotenv to automatically load environment variables set in the .env file present
+# in the application root.
 #
+# This file can be used to set variables like KAMAL_REGISTRY_PASSWORD or database passwords.
+# But for this reason you must ensure that .env files are not checked into Git or included
+# in your Dockerfile! The format is just key-value like:
 # ```
-# KAMAL_REGISTRY_PASSWORD=$KAMAL_REGISTRY_PASSWORD
-# RAILS_MASTER_KEY=$(cat config/master.key)
+# KAMAL_REGISTRY_PASSWORD=pw
+# DB_PASSWORD=secret123
 # ```
 #
-# You can also use [secret helpers](../commands/secrets) for some common password managers.
-# ```
-# SECRETS=$(kamal secrets fetch ...)
-#
-# REGISTRY_PASSWORD=$(kamal secrets extract REGISTRY_PASSWORD $SECRETS)
-# DB_PASSWORD=$(kamal secrets extract DB_PASSWORD $SECRETS)
-# ```
-#
-# If you store secrets directly in .kamal/secrets, ensure that it is not checked into version control.
-#
 # To pass the secrets you should list them under the `secret` key. When you do this the
 # other variables need to be moved under the `clear` key.
 #

lib/kamal/configuration/docs/healthcheck.yml

@@ -0,0 +1,59 @@
+# Healthcheck configuration
+#
+# On roles that are running Traefik, Kamal will supply a default healthcheck to `docker run`.
+# For other roles, by default no healthcheck is supplied.
+#
+# If no healthcheck is supplied and the image does not define one, then we wait for the container
+# to reach a running state and then pause for the readiness delay.
+#
+# The default healthcheck is `curl -f http://localhost:<port>/<path>`, so it assumes that `curl`
+# is available within the container.
+
+# Healthcheck options
+#
+# These go under the `healthcheck` key in the root or role configuration.
+healthcheck:
+
+  # Command
+  #
+  # The command to run, defaults to `curl -f http://localhost:<port>/<path>` on roles running Traefik
+  cmd: "curl -f http://localhost"
+
+  # Interval
+  #
+  # The Docker healthcheck interval, defaults to `1s`
+  interval: 10s
+
+  # Max attempts
+  #
+  # The maximum number of times we poll the container to see if it is healthy, defaults to `7`
+  # Each check is separated by an increasing interval starting with 1 second.
+  max_attempts: 3
+
+  # Port
+  #
+  # The port to use in the healthcheck, defaults to `3000`
+  port: "80"
+
+  # Path
+  #
+  # The path to use in the healthcheck, defaults to `/up`
+  path: /health
+
+  # Cords for zero-downtime deployments
+  #
+  # The cord file is used for zero-downtime deployments. The healthcheck is augmented with a check
+  # for the existance of the file. This allows us to delete the file and force the container to
+  # become unhealthy, causing Traefik to stop routing traffic to it.
+  #
+  # Kamal mounts a volume at this location and creates the file before starting the container.
+  # You can set the value to `false` to disable the cord file, but this loses the zero-downtime
+  # guarantee.
+  #
+  # The default value is `/tmp/kamal-cord`
+  cord: /cord
+
+  # Log lines
+  #
+  # Number of lines to log from the container when the healthcheck fails, defaults to `50`
+  log_lines: 100

lib/kamal/configuration/docs/proxy.yml

@@ -1,49 +1,79 @@
 # Proxy
 #
-# Kamal uses [kamal-proxy](https://github.com/basecamp/kamal-proxy) to provide
-# gapless deployments. It runs on ports 80 and 443 and forwards requests to the
-# application container.
+# **Experimental** [kamal-proxy](http://github.com/basecamp/kamal-proxy) is a
+# custom built specifically for Kamal. It will replace Traefik in Kamal v2.0,
+# but currently is available as an experimental feature.
 #
-# The proxy is configured in the root configuration under `proxy`. These are
+# When this is enabled, the proxy will be started on the hosts listed under the hosts key.
+# In addition, the kamal traefik command will be disabled and replaced by kamal proxy.
+#
+# The kamal proxy command works identically to kamal traefik on hosts that have not
+# been included. It will also handle switching between Traefik and kamal-proxy when you
+# run kamal proxy reboot.
+
+# Limitations
+#
+# Currently the proxy will run on ports 80 and 443 and will bind to those
+# ports on the host.
+#
+# There is no way to set custom options for `docker run` when booting the proxy.
+#
+# If you have custom Traefik configuration via labels or boot arguments they may
+# not have an equivalent in kamal-proxy.
+
+# Proxy settings
+#
+# The proxy is configured in the root configuration under `traefik`. These are
 # options that are set when deploying the application, not when booting the proxy
 #
 # They are application specific, so are not shared when multiple applications
-# run on the same proxy.
-#
-# The proxy is enabled by default on the primary role, but can be disabled by
-# setting `proxy: false`.
-#
-# It is disabled by default on all other roles, but can be enabled by setting
-# `proxy: true`, or providing a proxy configuration.
+# with the same proxy.
 proxy:
 
+  # Enabled
+  #
+  # Whether to enable experimental proxy support. Defaults to false
+  enabled: true
+
+  # Hosts
+  #
+  # The hosts to run the proxy on, instead of Traefik
+  # This is a temporary setting and will be removed when we full switch to kamal-proxy
+  #
+  # If you run `kamal traefik reboot`, then the proxy will be started on these hosts
+  # in place of traefik.
+  hosts:
+    - 10.0.0.1
+    - 10.0.0.2
+
   # Host
   #
-  # The hosts that will be used to serve the app. The proxy will only route requests
-  # to this host to your app.
+  # This is the host that will be used to serve the app. By setting this you can run
+  # multiple apps on the same server sharing the same instance of the proxy.
   #
-  # If no hosts are set, then all requests will be forwarded, except for matching
-  # requests for other apps deployed on that server that do have a host set.
+  # If this is set only requests that match this host will be forwarded by the proxy.
+  # if this is not set, then all requests will be forwarded, except for matching
+  # requests for other apps that do have a host set.
   host: foo.example.com
 
   # App port
   #
   # The port the application container is exposed on
-  #
   # Defaults to 80
   app_port: 3000
 
   # SSL
   #
-  # kamal-proxy can provide automatic HTTPS for your application via Let's Encrypt.
-  #
-  # This requires that we are deploying to a one server and the host option is set.
-  # The host value must point to the server we are deploying to and port 443 must be
-  # open for the Let's Encrypt challenge to succeed.
-  #
-  # Defaults to false
+  # Kamal Proxy can automatically obtain and renew TLS certificates for your applications.
+  # To ensure this set, the ssl flag. This only works if we are deploying to one server and
+  # the host flag is set.
   ssl: true
 
+  # Deploy timeout
+  #
+  # How long to wait for the app to boot when deploying, defaults to 30 seconds
+  deploy_timeout: 10s
+
   # Response timeout
   #
   # How long to wait for requests to complete before timing out, defaults to 30 seconds
@@ -80,7 +110,7 @@ proxy:
   #
   # Configure request logging for the proxy
   # You can specify request and response headers to log.
-  # By default, Cache-Control, Last-Modified and User-Agent request headers are logged
+  # By default, Cache-Control and Last-Modified request headers are logged
   logging:
     request_headers:
       - Cache-Control
@@ -91,10 +121,7 @@ proxy:
 
   # Forward headers
   #
-  # Whether to forward the X-Forwarded-For and X-Forwarded-Proto headers.
+  # Whether to forward the X-Forwarded-For and X-Forwarded-Proto headers (defaults to false)
   #
   # If you are behind a trusted proxy, you can set this to true to forward the headers.
-  #
-  # By default kamal-proxy will not forward the headers the ssl option is set to true, and
-  # will forward them if it is set to false.
   forward_headers: true

lib/kamal/configuration/docs/registry.yml

@@ -27,13 +27,11 @@ registry:
 # and [set up roles and permissions](https://cloud.google.com/artifact-registry/docs/access-control#permissions).
 # Normally, assigning a roles/artifactregistry.writer role should be sufficient.
 #
-# Once the service account is ready, you need to generate and download a JSON key and base64 encode it:
+# Once the service account is ready, you need to generate and download a JSON key, base64 encode it and add to .env:
 #
 # ```shell
-# base64 -i /path/to/key.json | tr -d "\\n")
+# echo "KAMAL_REGISTRY_PASSWORD=$(base64 -i /path/to/key.json)" | tr -d "\\n"  >> .env
 # ```
-# You'll then need to set the KAMAL_REGISTRY_PASSWORD secret to that value.
-#
 # Use the env variable as password along with _json_key_base64 as username.
 # Here’s the final configuration:
 

lib/kamal/configuration/docs/role.yml

@@ -26,12 +26,8 @@ servers:
   #
   # When there are other options to set, the list of hosts goes under the `hosts` key
   #
-  # By default only the primary role uses a proxy.
-  #
-  # For other roles, you can set it to `proxy: true` enable it and inherit the root proxy
-  # configuration or provide a map of options to override the root configuration.
-  #
-  # For the primary role, you can set `proxy: false` to disable the proxy.
+  # By default only the primary role uses Traefik, but you can set `traefik` to change
+  # it.
   #
   # You can also set a custom cmd to run in the container, and overwrite other settings
   # from the root configuration.
@@ -39,16 +35,18 @@ servers:
     hosts:
       - 172.1.0.3
       - 172.1.0.4: experiment1
+    traefik: true
     cmd: "bin/jobs"
     options:
       memory: 2g
       cpus: 4
-    logging:
+    healthcheck:
       ...
-    proxy:
+    logging:
       ...
     labels:
       my-label: workers
     env:
       ...
     asset_path: /public
+

lib/kamal/configuration/docs/traefik.yml

@@ -0,0 +1,62 @@
+# Traefik
+#
+# Traefik is a reverse proxy, used by Kamal for zero-downtime deployments.
+#
+# We start an instance on the hosts in it's own container.
+#
+# During a deployment:
+# 1. We start a new container which Traefik automatically detects due to the labels we have applied
+# 2. Traefik starts routing traffic to the new container
+# 3. We force the old container to fail it's healthcheck, causing Traefik to stop routing traffic to it
+# 4. We stop the old container
+
+# Traefik settings
+#
+# Traekik is configured in the root configuration under `traefik`.
+traefik:
+
+  # Image
+  #
+  # The Traefik image to use, defaults to `traefik:v2.10`
+  image: traefik:v2.9
+
+  # Host port
+  #
+  # The host port to publish the Traefik container on, defaults to `80`
+  host_port: "8080"
+
+  # Disabling publishing
+  #
+  # To avoid publishing the Traefik container, set this to `false`
+  publish: false
+
+  # Labels
+  #
+  # Additional labels to apply to the Traefik container
+  labels:
+    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"
+
+  # Arguments
+  #
+  # Additional arguments to pass to the Traefik container
+  args:
+    entryPoints.http.address: ":80"
+    entryPoints.http.forwardedHeaders.insecure: true
+    accesslog: true
+    accesslog.format: json
+
+  # Options
+  #
+  # Additional options to pass to `docker run`
+  options:
+    cpus: 2
+
+  # Environment variables
+  #
+  # See kamal docs env
+  env:
+    ...

lib/kamal/configuration/healthcheck.rb

@@ -0,0 +1,63 @@
+class Kamal::Configuration::Healthcheck
+  include Kamal::Configuration::Validation
+
+  attr_reader :healthcheck_config
+
+  def initialize(healthcheck_config:, context: "healthcheck")
+    @healthcheck_config = healthcheck_config || {}
+    validate! @healthcheck_config, context: context
+  end
+
+  def merge(other)
+    self.class.new healthcheck_config: healthcheck_config.deep_merge(other.healthcheck_config)
+  end
+
+  def cmd
+    healthcheck_config.fetch("cmd", http_health_check)
+  end
+
+  def port
+    healthcheck_config.fetch("port", 3000)
+  end
+
+  def path
+    healthcheck_config.fetch("path", "/up")
+  end
+
+  def max_attempts
+    healthcheck_config.fetch("max_attempts", 7)
+  end
+
+  def interval
+    healthcheck_config.fetch("interval", "1s")
+  end
+
+  def cord
+    healthcheck_config.fetch("cord", "/tmp/kamal-cord")
+  end
+
+  def log_lines
+    healthcheck_config.fetch("log_lines", 50)
+  end
+
+  def set_port_or_path?
+    healthcheck_config["port"].present? || healthcheck_config["path"].present?
+  end
+
+  def to_h
+    {
+      "cmd" => cmd,
+      "interval" => interval,
+      "max_attempts" => max_attempts,
+      "port" => port,
+      "path" => path,
+      "cord" => cord,
+      "log_lines" => log_lines
+    }
+  end
+
+  private
+    def http_health_check
+      "curl -f #{URI.join("http://localhost:#{port}", path)} || exit 1" if path.present? || port.present?
+    end
+end

lib/kamal/configuration/proxy.rb

@@ -1,41 +1,61 @@
 class Kamal::Configuration::Proxy
   include Kamal::Configuration::Validation
 
-  DEFAULT_LOG_REQUEST_HEADERS = [ "Cache-Control", "Last-Modified", "User-Agent" ]
-  CONTAINER_NAME = "kamal-proxy"
+  DEFAULT_HTTP_PORT = 80
+  DEFAULT_HTTPS_PORT = 443
+  DEFAULT_IMAGE = "basecamp/kamal-proxy:latest"
+  DEFAULT_LOG_REQUEST_HEADERS = [ "Cache-Control", "Last-Modified" ]
 
   delegate :argumentize, :optionize, to: Kamal::Utils
 
-  attr_reader :config, :proxy_config
-
-  def initialize(config:, proxy_config:, context: "proxy")
+  def initialize(config:)
     @config = config
-    @proxy_config = proxy_config
-    validate! @proxy_config, with: Kamal::Configuration::Validator::Proxy, context: context
+    @proxy_config = config.raw_config.proxy || {}
+    validate! proxy_config, with: Kamal::Configuration::Validator::Proxy
+  end
+
+  def enabled?
+    !!proxy_config.fetch("enabled", false)
+  end
+
+  def hosts
+    if enabled?
+      proxy_config.fetch("hosts", [])
+    else
+      []
+    end
   end
 
   def app_port
     proxy_config.fetch("app_port", 80)
   end
 
-  def ssl?
-    proxy_config.fetch("ssl", false)
+  def image
+    proxy_config.fetch("image", DEFAULT_IMAGE)
   end
 
-  def host
-    proxy_config["host"]
+  def container_name
+    "kamal-proxy"
+  end
+
+  def publish_args
+    argumentize "--publish", [ "#{DEFAULT_HTTP_PORT}:#{DEFAULT_HTTP_PORT}", "#{DEFAULT_HTTPS_PORT}:#{DEFAULT_HTTPS_PORT}" ]
+  end
+
+  def ssl?
+    proxy_config.fetch("ssl", false)
   end
 
   def deploy_options
     {
       host: proxy_config["host"],
       tls: proxy_config["ssl"],
-      "deploy-timeout": seconds_duration(config.deploy_timeout),
-      "drain-timeout": seconds_duration(config.drain_timeout),
-      "health-check-interval": seconds_duration(proxy_config.dig("healthcheck", "interval")),
-      "health-check-timeout": seconds_duration(proxy_config.dig("healthcheck", "timeout")),
-      "health-check-path": proxy_config.dig("healthcheck", "path"),
-      "target-timeout": seconds_duration(proxy_config["response_timeout"]),
+      "deploy-timeout": proxy_config["deploy_timeout"],
+      "drain-timeout": proxy_config["drain_timeout"],
+      "health-check-interval": proxy_config.dig("health_check", "interval"),
+      "health-check-timeout": proxy_config.dig("health_check", "timeout"),
+      "health-check-path": proxy_config.dig("health_check", "path"),
+      "target-timeout": proxy_config["response_timeout"],
       "buffer-requests": proxy_config.fetch("buffering", { "requests": true }).fetch("requests", true),
       "buffer-responses": proxy_config.fetch("buffering", { "responses": true }).fetch("responses", true),
       "buffer-memory": proxy_config.dig("buffering", "memory"),
@@ -47,20 +67,14 @@ class Kamal::Configuration::Proxy
     }.compact
   end
 
-  def deploy_command_args(target:)
-    optionize ({ target: "#{target}:#{app_port}" }).merge(deploy_options)
+  def deploy_command_args
+    optionize deploy_options
   end
 
-  def remove_command_args(target:)
-    optionize({ target: "#{target}:#{app_port}" })
-  end
-
-  def merge(other)
-    self.class.new config: config, proxy_config: proxy_config.deep_merge(other.proxy_config)
+  def config_directory_as_docker_volume
+    File.join config.run_directory_as_docker_volume, "proxy", "config"
   end
 
   private
-    def seconds_duration(value)
-      value ? "#{value}s" : nil
-    end
+    attr_reader :config, :proxy_config
 end

lib/kamal/configuration/registry.rb

@@ -21,14 +21,6 @@ class Kamal::Configuration::Registry
     lookup("password")
   end
 
-  def local?
-    server&.match?("^localhost[:$]")
-  end
-
-  def local_port
-    local? ? (server.split(":").last.to_i || 80) : nil
-  end
-
   private
     def lookup(key)
       if registry_config[key].is_a?(Array)

lib/kamal/configuration/role.rb

@@ -1,9 +1,10 @@
 class Kamal::Configuration::Role
   include Kamal::Configuration::Validation
 
+  CORD_FILE = "cord"
   delegate :argumentize, :optionize, to: Kamal::Utils
 
-  attr_reader :name, :config, :specialized_env, :specialized_logging, :specialized_proxy
+  attr_reader :name, :config, :specialized_env, :specialized_logging, :specialized_healthcheck
 
   alias to_s name
 
@@ -24,7 +25,9 @@ class Kamal::Configuration::Role
       logging_config: specializations.fetch("logging", {}),
       context: "servers/#{name}/logging"
 
-    initialize_specialized_proxy
+    @specialized_healthcheck = Kamal::Configuration::Healthcheck.new \
+      healthcheck_config: specializations.fetch("healthcheck", {}),
+      context: "servers/#{name}/healthcheck"
   end
 
   def primary_host
@@ -52,6 +55,10 @@ class Kamal::Configuration::Role
   end
 
   def labels
+    default_labels.merge(traefik_labels).merge(custom_labels)
+  end
+
+  def labels_for_proxy
     default_labels.merge(custom_labels)
   end
 
@@ -59,6 +66,10 @@ class Kamal::Configuration::Role
     argumentize "--label", labels
   end
 
+  def label_args_for_proxy
+    argumentize "--label", labels_for_proxy
+  end
+
   def logging_args
     logging.args
   end
@@ -67,24 +78,6 @@ class Kamal::Configuration::Role
     @logging ||= config.logging.merge(specialized_logging)
   end
 
-  def proxy
-    @proxy ||= config.proxy.merge(specialized_proxy) if running_proxy?
-  end
-
-  def running_proxy?
-    @running_proxy
-  end
-
-  def ssl?
-    running_proxy? && proxy.ssl?
-  end
-
-  def stop_args
-    # When deploying with the proxy, kamal-proxy will drain request before returning so we don't need to wait.
-    timeout = running_proxy? ? nil : config.drain_timeout
-
-    [ *argumentize("-t", timeout) ]
-  end
 
   def env(host)
     @envs ||= {}
@@ -104,7 +97,7 @@ class Kamal::Configuration::Role
   end
 
   def secrets_path
-    File.join(config.env_directory, "roles", "#{name}.env")
+    File.join(config.env_directory, "roles", "#{container_prefix}.env")
   end
 
   def asset_volume_args
@@ -112,8 +105,72 @@ class Kamal::Configuration::Role
   end
 
 
+  def health_check_args(cord: true)
+    if running_traefik? || healthcheck.set_port_or_path?
+      if cord && uses_cord?
+        optionize({ "health-cmd" => health_check_cmd_with_cord, "health-interval" => healthcheck.interval })
+          .concat(cord_volume.docker_args)
+      else
+        optionize({ "health-cmd" => healthcheck.cmd, "health-interval" => healthcheck.interval })
+      end
+    else
+      []
+    end
+  end
+
+  def healthcheck
+    @healthcheck ||=
+      if running_traefik?
+        config.healthcheck.merge(specialized_healthcheck)
+      else
+        specialized_healthcheck
+      end
+  end
+
+  def health_check_cmd_with_cord
+    "(#{healthcheck.cmd}) && (stat #{cord_container_file} > /dev/null || exit 1)"
+  end
+
+
+  def running_traefik?
+    if specializations["traefik"].nil?
+      primary?
+    else
+      specializations["traefik"]
+    end
+  end
+
   def primary?
-    name == @config.primary_role_name
+    self == @config.primary_role
+  end
+
+
+  def uses_cord?
+    running_traefik? && cord_volume && healthcheck.cmd.present?
+  end
+
+  def cord_host_directory
+    File.join config.run_directory_as_docker_volume, "cords", [ container_prefix, config.run_id ].join("-")
+  end
+
+  def cord_volume
+    if (cord = healthcheck.cord)
+      @cord_volume ||= Kamal::Configuration::Volume.new \
+        host_path: File.join(config.run_directory, "cords", [ container_prefix, config.run_id ].join("-")),
+        container_path: cord
+    end
+  end
+
+  def cord_host_file
+    File.join cord_volume.host_path, CORD_FILE
+  end
+
+  def cord_container_directory
+    health_check_options.fetch("cord", nil)
+  end
+
+  def cord_container_file
+    File.join cord_volume.container_path, CORD_FILE
   end
 
 
@@ -131,52 +188,25 @@ class Kamal::Configuration::Role
   end
 
   def assets?
-    asset_path.present? && running_proxy?
+    asset_path.present? && running_traefik?
   end
 
-  def asset_volume(version = config.version)
+  def asset_volume(version = nil)
     if assets?
       Kamal::Configuration::Volume.new \
-        host_path: asset_volume_directory(version), container_path: asset_path
+        host_path: asset_volume_path(version), container_path: asset_path
     end
   end
 
-  def asset_extracted_directory(version = config.version)
-    File.join config.assets_directory, "extracted", [ name, version ].join("-")
+  def asset_extracted_path(version = nil)
+    File.join config.run_directory, "assets", "extracted", container_name(version)
   end
 
-  def asset_volume_directory(version = config.version)
-    File.join config.assets_directory, "volumes", [ name, version ].join("-")
-  end
-
-  def ensure_one_host_for_ssl
-    if running_proxy? && proxy.ssl? && hosts.size > 1
-      raise Kamal::ConfigurationError, "SSL is only supported on a single server, found #{hosts.size} servers for role #{name}"
-    end
+  def asset_volume_path(version = nil)
+    File.join config.run_directory, "assets", "volumes", container_name(version)
   end
 
   private
-    def initialize_specialized_proxy
-      proxy_specializations = specializations["proxy"]
-
-      if primary?
-        # only false means no proxy for non-primary roles
-        @running_proxy = proxy_specializations != false
-      else
-        # false and nil both mean no proxy for non-primary roles
-        @running_proxy = !!proxy_specializations
-      end
-
-      if running_proxy?
-        proxy_config = proxy_specializations == true || proxy_specializations.nil? ? {} : proxy_specializations
-
-        @specialized_proxy = Kamal::Configuration::Proxy.new \
-          config: config,
-          proxy_config: proxy_config,
-          context: "servers/#{name}/proxy"
-      end
-    end
-
     def tagged_hosts
       {}.tap do |tagged_hosts|
         extract_hosts_from_config.map do |host_config|
@@ -211,6 +241,27 @@ class Kamal::Configuration::Role
       end
     end
 
+    def traefik_labels
+      if running_traefik?
+        {
+          # Setting a service property ensures that the generated service name will be consistent between versions
+          "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"
+        }
+      else
+        {}
+      end
+    end
+
+    def traefik_service
+      container_prefix
+    end
+
     def custom_labels
       Hash.new.tap do |labels|
         labels.merge!(config.labels) if config.labels.present?

lib/kamal/configuration/traefik.rb

@@ -0,0 +1,78 @@
+class Kamal::Configuration::Traefik
+  delegate :argumentize, to: Kamal::Utils
+
+  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"
+  }
+
+  include Kamal::Configuration::Validation
+
+  attr_reader :config, :traefik_config
+
+  def initialize(config:)
+    @config = config
+    @traefik_config = config.raw_config.traefik || {}
+    validate! traefik_config
+  end
+
+  def publish?
+    traefik_config["publish"] != false
+  end
+
+  def labels
+    DEFAULT_LABELS.merge(traefik_config["labels"] || {})
+  end
+
+  def env
+    Kamal::Configuration::Env.new \
+      config: traefik_config.fetch("env", {}),
+      secrets: config.secrets,
+      context: "traefik/env"
+  end
+
+  def host_port
+    traefik_config.fetch("host_port", CONTAINER_PORT)
+  end
+
+  def options
+    traefik_config.fetch("options", {})
+  end
+
+  def port
+    "#{host_port}:#{CONTAINER_PORT}"
+  end
+
+  def args
+    DEFAULT_ARGS.merge(traefik_config.fetch("args", {}))
+  end
+
+  def image
+    traefik_config.fetch("image", DEFAULT_IMAGE)
+  end
+
+  def env_args
+    [ *env.clear_args, *argumentize("--env-file", secrets_path) ]
+  end
+
+  def env_directory
+    File.join(config.env_directory, "traefik")
+  end
+
+  def secrets_io
+    env.secrets_io
+  end
+
+  def secrets_path
+    File.join(config.env_directory, "traefik", "traefik.env")
+  end
+end

lib/kamal/configuration/validator.rb

@@ -24,9 +24,7 @@ class Kamal::Configuration::Validator
             example_value = example[key]
 
             if example_value == "..."
-              unless key.to_s == "proxy" && boolean?(value.class)
-                validate_type! value, *(Array if key == :servers), Hash
-              end
+              validate_type! value, *(Array if key == :servers), Hash
             elsif key == "hosts"
               validate_servers! value
             elsif example_value.is_a?(Array)

lib/kamal/configuration/validator/proxy.rb

@@ -1,11 +1,9 @@
 class Kamal::Configuration::Validator::Proxy < Kamal::Configuration::Validator
   def validate!
-    unless config.nil?
-      super
+    super
 
-      if config["host"].blank? && config["ssl"]
-        error "Must set a host to enable automatic SSL"
-      end
+    if config["host"].blank? && config["ssl"]
+      error "Must set a host to enable automatic SSL"
     end
   end
 end

lib/kamal/configuration/validator/registry.rb

@@ -15,12 +15,10 @@ class Kamal::Configuration::Validator::Registry < Kamal::Configuration::Validato
       with_context(key) do
         value = config[key]
 
-        unless config["server"]&.match?("^localhost[:$]")
-          error "is required" unless value.present?
+        error "is required" unless value.present?
 
-          unless value.is_a?(String) || (value.is_a?(Array) && value.size == 1 && value.first.is_a?(String))
-            error "should be a string or an array with one string (for secret lookup)"
-          end
+        unless value.is_a?(String) || (value.is_a?(Array) && value.size == 1 && value.first.is_a?(String))
+          error "should be a string or an array with one string (for secret lookup)"
         end
       end
     end

lib/kamal/secrets.rb

@@ -8,14 +8,10 @@ class Kamal::Secrets
   def initialize(destination: nil)
     @secrets_files = \
       [ ".kamal/secrets-common", ".kamal/secrets#{(".#{destination}" if destination)}" ].select { |f| File.exist?(f) }
-    @mutex = Mutex.new
   end
 
   def [](key)
-    # Fetching secrets may ask the user for input, so ensure only one thread does that
-    @mutex.synchronize do
-      secrets.fetch(key)
-    end
+    secrets.fetch(key)
   rescue KeyError
     if secrets_files
       raise Kamal::ConfigurationError, "Secret '#{key}' not found in #{secrets_files.join(", ")}"

lib/kamal/utils.rb

@@ -101,8 +101,4 @@ module Kamal::Utils
       arch
     end
   end
-
-  def older_version?(version, other_version)
-    Gem::Version.new(version.delete_prefix("v")) < Gem::Version.new(other_version.delete_prefix("v"))
-  end
 end

lib/kamal/version.rb

@@ -1,3 +1,3 @@
 module Kamal
-  VERSION = "2.0.0.rc2"
+  VERSION = "2.0.0.alpha"
 end

test/cli/accessory_test.rb

@@ -15,7 +15,7 @@ class CliAccessoryTest < CliTestCase
 
     run_command("boot", "mysql").tap do |output|
       assert_match /docker login.*on 1.1.1.3/, output
-      assert_match "docker run --name app-mysql --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --publish 3306:3306 --env MYSQL_ROOT_HOST=\"%\" --env-file .kamal/apps/app/env/accessories/mysql.env --volume $PWD/app-mysql/etc/mysql/my.cnf:/etc/mysql/my.cnf --volume $PWD/app-mysql/data:/var/lib/mysql --label service=\"app-mysql\" mysql:5.7 on 1.1.1.3", output
+      assert_match "docker run --name app-mysql --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --publish 3306:3306 --env MYSQL_ROOT_HOST=\"%\" --env-file .kamal/env/accessories/app-mysql.env --volume $PWD/app-mysql/etc/mysql/my.cnf:/etc/mysql/my.cnf --volume $PWD/app-mysql/data:/var/lib/mysql --label service=\"app-mysql\" mysql:5.7 on 1.1.1.3", output
     end
   end
 
@@ -32,9 +32,9 @@ class CliAccessoryTest < CliTestCase
       assert_match /docker network create kamal.*on 1.1.1.1/, output
       assert_match /docker network create kamal.*on 1.1.1.2/, output
       assert_match /docker network create kamal.*on 1.1.1.3/, output
-      assert_match "docker run --name app-mysql --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --publish 3306:3306 --env MYSQL_ROOT_HOST=\"%\" --env-file .kamal/apps/app/env/accessories/mysql.env --volume $PWD/app-mysql/etc/mysql/my.cnf:/etc/mysql/my.cnf --volume $PWD/app-mysql/data:/var/lib/mysql --label service=\"app-mysql\" mysql:5.7 on 1.1.1.3", output
-      assert_match "docker run --name app-redis --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --publish 6379:6379 --env-file .kamal/apps/app/env/accessories/redis.env --volume $PWD/app-redis/data:/data --label service=\"app-redis\" redis:latest on 1.1.1.1", output
-      assert_match "docker run --name app-redis --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --publish 6379:6379 --env-file .kamal/apps/app/env/accessories/redis.env --volume $PWD/app-redis/data:/data --label service=\"app-redis\" redis:latest on 1.1.1.2", output
+      assert_match "docker run --name app-mysql --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --publish 3306:3306 --env MYSQL_ROOT_HOST=\"%\" --env-file .kamal/env/accessories/app-mysql.env --volume $PWD/app-mysql/etc/mysql/my.cnf:/etc/mysql/my.cnf --volume $PWD/app-mysql/data:/var/lib/mysql --label service=\"app-mysql\" mysql:5.7 on 1.1.1.3", output
+      assert_match "docker run --name app-redis --detach --restart unless-stopped --network kamal --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
+      assert_match "docker run --name app-redis --detach --restart unless-stopped --network kamal --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
 
@@ -203,8 +203,8 @@ class CliAccessoryTest < CliTestCase
     run_command("boot", "redis", "--hosts", "1.1.1.1").tap do |output|
       assert_match /docker login.*on 1.1.1.1/, output
       assert_no_match /docker login.*on 1.1.1.2/, output
-      assert_match "docker run --name app-redis --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --publish 6379:6379 --env-file .kamal/apps/app/env/accessories/redis.env --volume $PWD/app-redis/data:/data --label service=\"app-redis\" redis:latest on 1.1.1.1", output
-      assert_no_match "docker run --name app-redis --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --publish 6379:6379 --env-file .kamal/apps/app/env/accessories/redis.env --volume $PWD/app-redis/data:/data --label service=\"app-redis\" redis:latest on 1.1.1.2", output
+      assert_match "docker run --name app-redis --detach --restart unless-stopped --network kamal --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
+      assert_no_match "docker run --name app-redis --detach --restart unless-stopped --network kamal --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
 
@@ -215,32 +215,11 @@ class CliAccessoryTest < CliTestCase
     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
       assert_no_match /docker login.*on 1.1.1.3/, output
-      assert_match "docker run --name app-redis --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --publish 6379:6379 --env-file .kamal/apps/app/env/accessories/redis.env --volume $PWD/app-redis/data:/data --label service=\"app-redis\" redis:latest on 1.1.1.1", output
-      assert_no_match "docker run --name app-redis --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --publish 6379:6379 --env-file .kamal/apps/app/env/accessories/redis.env --volume $PWD/app-redis/data:/data --label service=\"app-redis\" redis:latest on 1.1.1.3", output
+      assert_match "docker run --name app-redis --detach --restart unless-stopped --network kamal --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
+      assert_no_match "docker run --name app-redis --detach --restart unless-stopped --network kamal --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
 
-  test "upgrade" do
-    run_command("upgrade", "-y", "all").tap do |output|
-      assert_match "Upgrading all accessories on 1.1.1.3,1.1.1.1,1.1.1.2...", output
-      assert_match "docker network create kamal on 1.1.1.3", output
-      assert_match "docker container stop app-mysql on 1.1.1.3", output
-      assert_match "docker run --name app-mysql --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --publish 3306:3306 --env MYSQL_ROOT_HOST="%" --env-file .kamal/apps/app/env/accessories/mysql.env --volume $PWD/app-mysql/etc/mysql/my.cnf:/etc/mysql/my.cnf --volume $PWD/app-mysql/data:/var/lib/mysql --label service=\"app-mysql\" mysql:5.7 on 1.1.1.3", output
-      assert_match "Upgraded all accessories on 1.1.1.3,1.1.1.1,1.1.1.2...", output
-    end
-  end
-
-  test "upgrade rolling" do
-    run_command("upgrade", "--rolling", "-y", "all").tap do |output|
-      assert_match "Upgrading all accessories on 1.1.1.3...", output
-      assert_match "docker network create kamal on 1.1.1.3", output
-      assert_match "docker container stop app-mysql on 1.1.1.3", output
-      assert_match "docker run --name app-mysql --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --publish 3306:3306 --env MYSQL_ROOT_HOST="%" --env-file .kamal/apps/app/env/accessories/mysql.env --volume $PWD/app-mysql/etc/mysql/my.cnf:/etc/mysql/my.cnf --volume $PWD/app-mysql/data:/var/lib/mysql --label service=\"app-mysql\" mysql:5.7 on 1.1.1.3", output
-      assert_match "Upgraded all accessories 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

@@ -5,7 +5,7 @@ class CliAppTest < CliTestCase
     stub_running
     run_command("boot").tap do |output|
       assert_match "docker tag dhh/app:latest dhh/app:latest", output
-      assert_match /docker run --detach --restart unless-stopped --name app-web-latest --network kamal --hostname 1.1.1.1-[0-9a-f]{12} /, output
+      assert_match /docker run --detach --restart unless-stopped --name app-web-latest --hostname 1.1.1.1-[0-9a-f]{12} /, output
       assert_match "docker container ls --all --filter name=^app-web-123$ --quiet | xargs docker stop", output
     end
   end
@@ -18,18 +18,26 @@ class CliAppTest < CliTestCase
       .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-latest$", "--quiet", raise_on_non_zero_exit: false)
       .returns("12345678") # running version
 
+    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
+      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-latest$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
+      .returns("running") # health check
+
     SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
       .with(:sh, "-c", "'docker ps --latest --format '\\''{{.Names}}'\\'' --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --format '\\''{{.Names}}'\\'' --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting'", "|", :head, "-1", "|", "while read line; do echo ${line#app-web-}; done", raise_on_non_zero_exit: false)
       .returns("123") # old version
 
     SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
-      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-latest$", "--quiet")
-      .returns("12345678") # running version
+      .with(:docker, :inspect, "-f '{{ range .Mounts }}{{printf \"%s %s\\n\" .Source .Destination}}{{ end }}'", "app-web-123", "|", :awk, "'$2 == \"/tmp/kamal-cord\" {print $1}'", raise_on_non_zero_exit: false)
+      .returns("cordfile") # old version
+
+    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
+      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-123$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
+      .returns("unhealthy") # old version unhealthy
 
     run_command("boot").tap do |output|
       assert_match /Renaming container .* to .* as already deployed on 1.1.1.1/, output # Rename
       assert_match /docker rename app-web-latest app-web-latest_replaced_[0-9a-f]{16}/, output
-      assert_match /docker run --detach --restart unless-stopped --name app-web-latest --network kamal --hostname 1.1.1.1-[0-9a-f]{12} /, output
+      assert_match /docker run --detach --restart unless-stopped --name app-web-latest --hostname 1.1.1.1-[0-9a-f]{12} /, output
       assert_match "docker container ls --all --filter name=^app-web-123$ --quiet | xargs docker stop", output
     end
   ensure
@@ -37,7 +45,7 @@ class CliAppTest < CliTestCase
   end
 
   test "boot uses group strategy when specified" do
-    Kamal::Cli::App.any_instance.stubs(:on).with("1.1.1.1").times(2) # ensure locks dir, acquire & release lock
+    Kamal::Cli::App.any_instance.stubs(:on).with("1.1.1.1").times(3) # ensure locks dir, acquire & release lock
     Kamal::Cli::App.any_instance.stubs(:on).with([ "1.1.1.1" ]) # tag container
 
     # Strategy is used when booting the containers
@@ -62,21 +70,25 @@ class CliAppTest < CliTestCase
       .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-latest$", "--quiet", raise_on_non_zero_exit: false)
       .returns("12345678") # running version
 
+    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
+      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-latest$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
+      .returns("running") # health check
+
     SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
       .with(:sh, "-c", "'docker ps --latest --format '\\''{{.Names}}'\\'' --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --format '\\''{{.Names}}'\\'' --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting'", "|", :head, "-1", "|", "while read line; do echo ${line#app-web-}; done", raise_on_non_zero_exit: false)
       .returns("123").twice # old version
 
     SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
-      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-latest$", "--quiet")
-      .returns("12345678") # running version
+      .with(:docker, :inspect, "-f '{{ range .Mounts }}{{printf \"%s %s\\n\" .Source .Destination}}{{ end }}'", "app-web-123", "|", :awk, "'$2 == \"/tmp/kamal-cord\" {print $1}'", raise_on_non_zero_exit: false)
+      .returns("") # old version
 
     run_command("boot", config: :with_assets).tap do |output|
       assert_match "docker tag dhh/app:latest dhh/app:latest", output
-      assert_match "/usr/bin/env mkdir -p .kamal/apps/app/assets/volumes/web-latest ; cp -rnT .kamal/apps/app/assets/extracted/web-latest .kamal/apps/app/assets/volumes/web-latest ; cp -rnT .kamal/apps/app/assets/extracted/web-latest .kamal/apps/app/assets/volumes/web-123 || true ; cp -rnT .kamal/apps/app/assets/extracted/web-123 .kamal/apps/app/assets/volumes/web-latest || true", output
-      assert_match "/usr/bin/env mkdir -p .kamal/apps/app/assets/extracted/web-latest && docker stop -t 1 app-web-assets 2> /dev/null || true && docker run --name app-web-assets --detach --rm --entrypoint sleep dhh/app:latest 1000000 && docker cp -L app-web-assets:/public/assets/. .kamal/apps/app/assets/extracted/web-latest && docker stop -t 1 app-web-assets", output
-      assert_match /docker run --detach --restart unless-stopped --name app-web-latest --network kamal --hostname 1.1.1.1-[0-9a-f]{12} /, output
+      assert_match "/usr/bin/env mkdir -p .kamal/assets/volumes/app-web-latest ; cp -rnT .kamal/assets/extracted/app-web-latest .kamal/assets/volumes/app-web-latest ; cp -rnT .kamal/assets/extracted/app-web-latest .kamal/assets/volumes/app-web-123 || true ; cp -rnT .kamal/assets/extracted/app-web-123 .kamal/assets/volumes/app-web-latest || true", output
+      assert_match "/usr/bin/env mkdir -p .kamal/assets/extracted/app-web-latest && docker stop -t 1 app-web-assets 2> /dev/null || true && docker run --name app-web-assets --detach --rm dhh/app:latest sleep 1000000 && docker cp -L app-web-assets:/public/assets/. .kamal/assets/extracted/app-web-latest && docker stop -t 1 app-web-assets", output
+      assert_match /docker run --detach --restart unless-stopped --name app-web-latest --hostname 1.1.1.1-[0-9a-f]{12} /, output
       assert_match "docker container ls --all --filter name=^app-web-123$ --quiet | xargs docker stop", output
-      assert_match "/usr/bin/env find .kamal/apps/app/assets/extracted -maxdepth 1 -name 'web-*' ! -name web-latest -exec rm -rf \"{}\" + ; find .kamal/apps/app/assets/volumes -maxdepth 1 -name 'web-*' ! -name web-latest -exec rm -rf \"{}\" +", output
+      assert_match "/usr/bin/env find .kamal/assets/extracted -maxdepth 1 -name 'app-web-*' ! -name app-web-latest -exec rm -rf \"{}\" + ; find .kamal/assets/volumes -maxdepth 1 -name 'app-web-*' ! -name app-web-latest -exec rm -rf \"{}\" +", output
     end
   end
 
@@ -84,20 +96,24 @@ class CliAppTest < CliTestCase
     Object.any_instance.stubs(:sleep)
 
     SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
-      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-latest$", "--quiet", raise_on_non_zero_exit: false)
-      .returns("12345678") # running version
+    .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-latest$", "--quiet", raise_on_non_zero_exit: false)
+    .returns("12345678") # running version
 
     SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
-      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-latest$", "--quiet")
-      .returns("12345678") # running version
+      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-latest$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
+      .returns("running") # health check
 
     SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
       .with(:sh, "-c", "'docker ps --latest --format '\\''{{.Names}}'\\'' --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --format '\\''{{.Names}}'\\'' --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting'", "|", :head, "-1", "|", "while read line; do echo ${line#app-web-}; done", raise_on_non_zero_exit: false)
       .returns("123") # old version
 
+    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
+      .with(:docker, :inspect, "-f '{{ range .Mounts }}{{printf \"%s %s\\n\" .Source .Destination}}{{ end }}'", "app-web-123", "|", :awk, "'$2 == \"/tmp/kamal-cord\" {print $1}'", raise_on_non_zero_exit: false)
+      .returns("") # old version
+
     run_command("boot", config: :with_env_tags).tap do |output|
       assert_match "docker tag dhh/app:latest dhh/app:latest", output
-      assert_match %r{docker run --detach --restart unless-stopped --name app-web-latest --network kamal --hostname 1.1.1.1-[0-9a-f]{12} -e KAMAL_CONTAINER_NAME="app-web-latest" -e KAMAL_VERSION="latest" --env TEST="root" --env EXPERIMENT="disabled" --env SITE="site1"}, output
+      assert_match %r{docker run --detach --restart unless-stopped --name app-web-latest --hostname 1.1.1.1-[0-9a-f]{12} -e KAMAL_CONTAINER_NAME="app-web-latest" -e KAMAL_VERSION="latest" --env TEST="root" --env EXPERIMENT="disabled" --env SITE="site1"}, output
       assert_match "docker container ls --all --filter name=^app-web-123$ --quiet | xargs docker stop", output
     end
   end
@@ -107,6 +123,14 @@ class CliAppTest < CliTestCase
 
     SSHKit::Backend::Abstract.any_instance.stubs(:capture_with_info).returns("123") # old version
 
+    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
+      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-latest$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
+      .returns("running").at_least_once # web health check passing
+
+    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
+      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-123$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
+      .returns("unhealthy").at_least_once # web health check failing
+
     SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
       .with(:docker, :container, :ls, "--all", "--filter", "name=^app-workers-latest$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
       .returns("running").at_least_once # workers health check
@@ -116,7 +140,7 @@ class CliAppTest < CliTestCase
       assert_match "Waiting for the first healthy web container before booting workers on 1.1.1.4...", output
       assert_match "First web container is healthy, booting workers on 1.1.1.3", output
       assert_match "First web container is healthy, booting workers on 1.1.1.4", output
-    end
+  end
   end
 
   test "boot with web barrier closed" do
@@ -126,11 +150,9 @@ class CliAppTest < CliTestCase
 
     SSHKit::Backend::Abstract.any_instance.stubs(:capture_with_info).returns("123") # old version
 
-    SSHKit::Backend::Abstract.any_instance.stubs(:execute).returns("")
-    SSHKit::Backend::Abstract.any_instance.stubs(:execute)
-      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-latest$", "--quiet", "|", :xargs, :docker, :stop, raise_on_non_zero_exit: false)
-    SSHKit::Backend::Abstract.any_instance.expects(:execute)
-      .with(:docker, :exec, "kamal-proxy", "kamal-proxy", :deploy, "app-web", "--target", "\"123:80\"", "--deploy-timeout", "\"1s\"", "--drain-timeout", "\"30s\"", "--buffer-requests", "--buffer-responses", "--log-request-header", "\"Cache-Control\"", "--log-request-header", "\"Last-Modified\"", "--log-request-header", "\"User-Agent\"").raises(SSHKit::Command::Failed.new("Failed to deploy"))
+    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
+      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-latest$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
+      .returns("unhealthy").at_least_once # web health check failing
 
     stderred do
       run_command("boot", config: :with_roles, host: nil, allow_execute_error: true).tap do |output|
@@ -138,59 +160,17 @@ class CliAppTest < CliTestCase
         assert_match "Waiting for the first healthy web container before booting workers on 1.1.1.4...", output
         assert_match "First web container is unhealthy, not booting workers on 1.1.1.3", output
         assert_match "First web container is unhealthy, not booting workers on 1.1.1.4", output
+        assert_match "Running docker container ls --all --filter name=^app-web-latest$ --quiet | xargs docker stop on 1.1.1.1", output
+        assert_match "Running docker container ls --all --filter name=^app-web-latest$ --quiet | xargs docker stop on 1.1.1.2", output
       end
     end
   ensure
     Thread.report_on_exception = true
   end
 
-  test "boot with worker errors" do
-    Thread.report_on_exception = false
-
-    Object.any_instance.stubs(:sleep)
-
-    SSHKit::Backend::Abstract.any_instance.stubs(:capture_with_info).returns("123") # old version
-
-    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
-      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-workers-latest$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
-      .returns("unhealthy").at_least_once # workers health check
-
-    run_command("boot", config: :with_roles, host: nil, allow_execute_error: true).tap do |output|
-      assert_match "Waiting for the first healthy web container before booting workers on 1.1.1.3...", output
-      assert_match "Waiting for the first healthy web container before booting workers on 1.1.1.4...", output
-      assert_match "First web container is healthy, booting workers on 1.1.1.3", output
-      assert_match "First web container is healthy, booting workers on 1.1.1.4", output
-      assert_match "ERROR Failed to boot workers on 1.1.1.3", output
-      assert_match "ERROR Failed to boot workers on 1.1.1.4", output
-    end
-  ensure
-    Thread.report_on_exception = true
-  end
-
-  test "boot with worker ready then not" do
-    Thread.report_on_exception = false
-
-    Object.any_instance.stubs(:sleep)
-
-    SSHKit::Backend::Abstract.any_instance.stubs(:capture_with_info).returns("123") # old version
-
-    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
-      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-workers-latest$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
-      .returns("running", "stopped").at_least_once # workers health check
-
-    run_command("boot", config: :with_roles, host: "1.1.1.3", allow_execute_error: true).tap do |output|
-      assert_match "ERROR Failed to boot workers on 1.1.1.3", output
-    end
-  ensure
-    Thread.report_on_exception = true
-  end
-
   test "start" do
-    SSHKit::Backend::Abstract.any_instance.stubs(:capture_with_info).returns("999") # old version
-
     run_command("start").tap do |output|
       assert_match "docker start app-web-999", output
-      assert_match "docker exec kamal-proxy kamal-proxy deploy app-web --target \"999:80\" --deploy-timeout \"30s\" --drain-timeout \"30s\" --buffer-requests --buffer-responses --log-request-header \"Cache-Control\" --log-request-header \"Last-Modified\"", output
     end
   end
 
@@ -263,13 +243,13 @@ class CliAppTest < CliTestCase
 
   test "exec" do
     run_command("exec", "ruby -v").tap do |output|
-      assert_match "docker run --rm --network kamal --env-file .kamal/apps/app/env/roles/web.env 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
 
   test "exec separate arguments" do
     run_command("exec", "ruby", " -v").tap do |output|
-      assert_match "docker run --rm --network kamal --env-file .kamal/apps/app/env/roles/web.env 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
 
@@ -282,7 +262,7 @@ class CliAppTest < CliTestCase
 
   test "exec interactive" do
     SSHKit::Backend::Abstract.any_instance.expects(:exec)
-      .with("ssh -t root@1.1.1.1 -p 22 'docker run -it --rm --network kamal --env-file .kamal/apps/app/env/roles/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
@@ -315,11 +295,11 @@ class CliAppTest < CliTestCase
     SSHKit::Backend::Abstract.any_instance.stubs(:exec)
       .with("ssh -t root@1.1.1.1 'sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1| xargs docker logs --timestamps --tail 10 2>&1'")
 
-    assert_match "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --timestamps --tail 100 2>&1", run_command("logs")
+    assert_match "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --tail 100 2>&1", run_command("logs")
 
-    assert_match "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --timestamps 2>&1 | grep 'hey'", run_command("logs", "--grep", "hey")
+    assert_match "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs 2>&1 | grep 'hey'", run_command("logs", "--grep", "hey")
 
-    assert_match "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --timestamps 2>&1 | grep 'hey' -C 2", run_command("logs", "--grep", "hey", "--grep-options", "-C 2")
+    assert_match "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs 2>&1 | grep 'hey' -C 2", run_command("logs", "--grep", "hey", "--grep-options", "-C 2")
   end
 
   test "logs with follow" do
@@ -362,7 +342,7 @@ class CliAppTest < CliTestCase
     hostname = "this-hostname-is-really-unacceptably-long-to-be-honest.example.com"
 
     stdouted { Kamal::Cli::App.start([ "boot", "-c", "test/fixtures/deploy_with_uncommon_hostnames.yml", "--hosts", hostname ]) }.tap do |output|
-      assert_match /docker run --detach --restart unless-stopped --name app-web-latest --network kamal --hostname this-hostname-is-really-unacceptably-long-to-be-hon-[0-9a-f]{12} /, output
+      assert_match /docker run --detach --restart unless-stopped --name app-web-latest --hostname this-hostname-is-really-unacceptably-long-to-be-hon-[0-9a-f]{12} /, output
     end
   end
 
@@ -372,7 +352,7 @@ class CliAppTest < CliTestCase
     hostname = "this-hostname-with-random-part-is-too-long.example.com"
 
     stdouted { Kamal::Cli::App.start([ "boot", "-c", "test/fixtures/deploy_with_uncommon_hostnames.yml", "--hosts", hostname ]) }.tap do |output|
-      assert_match /docker run --detach --restart unless-stopped --name app-web-latest --network kamal --hostname this-hostname-with-random-part-is-too-long.example-[0-9a-f]{12} /, output
+      assert_match /docker run --detach --restart unless-stopped --name app-web-latest --hostname this-hostname-with-random-part-is-too-long.example-[0-9a-f]{12} /, output
     end
   end
 
@@ -382,21 +362,12 @@ class CliAppTest < CliTestCase
     run_command("boot", config: :with_proxy).tap do |output|
       assert_match /Renaming container .* to .* as already deployed on 1.1.1.1/, output # Rename
       assert_match /docker rename app-web-latest app-web-latest_replaced_[0-9a-f]{16}/, output
-      assert_match /docker run --detach --restart unless-stopped --name app-web-latest --network kamal --hostname 1.1.1.1-[0-9a-f]{12} -e KAMAL_CONTAINER_NAME="app-web-latest" -e KAMAL_VERSION="latest" --env-file .kamal\/apps\/app\/env\/roles\/web.env --log-opt max-size="10m" --label service="app" --label role="web" --label destination dhh\/app:latest/, output
+      assert_match /docker run --detach --restart unless-stopped --name app-web-latest --network kamal --hostname 1.1.1.1-[0-9a-f]{12} -e KAMAL_CONTAINER_NAME="app-web-latest" -e KAMAL_VERSION="latest" --env-file .kamal\/env\/roles\/app-web.env --log-opt max-size="10m" --label service="app" --label role="web" --label destination dhh\/app:latest/, output
       assert_match /docker exec kamal-proxy kamal-proxy deploy app-web --target "123:80"/, output
       assert_match "docker container ls --all --filter name=^app-web-123$ --quiet | xargs docker stop", output
     end
   end
 
-  test "boot proxy with role specific config" do
-    SSHKit::Backend::Abstract.any_instance.stubs(:capture_with_info).returns("123") # old version
-
-    run_command("boot", config: :with_proxy_roles, host: nil).tap do |output|
-      assert_match "docker exec kamal-proxy kamal-proxy deploy app-web --target \"123:80\" --deploy-timeout \"6s\" --drain-timeout \"30s\" --target-timeout \"10s\" --buffer-requests --buffer-responses --log-request-header \"Cache-Control\" --log-request-header \"Last-Modified\" --log-request-header \"User-Agent\"", output
-      assert_match "docker exec kamal-proxy kamal-proxy deploy app-web2 --target \"123:80\" --deploy-timeout \"6s\" --drain-timeout \"30s\" --target-timeout \"15s\" --buffer-requests --buffer-responses --log-request-header \"Cache-Control\" --log-request-header \"Last-Modified\" --log-request-header \"User-Agent\"", output
-    end
-  end
-
   private
     def run_command(*command, config: :with_accessories, host: "1.1.1.1", allow_execute_error: false)
       stdouted do
@@ -410,5 +381,13 @@ class CliAppTest < CliTestCase
       Object.any_instance.stubs(:sleep)
 
       SSHKit::Backend::Abstract.any_instance.stubs(:capture_with_info).returns("123") # old version
+
+      SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
+        .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-latest$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
+        .returns("running") # health check
+
+      SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
+        .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-123$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
+        .returns("unhealthy") # health check
     end
 end

test/cli/build_test.rb

@@ -49,7 +49,7 @@ class CliBuildTest < CliTestCase
       SSHKit::Backend::Abstract.any_instance.expects(:execute).with(:git, "-C", build_directory, :submodule, :update, "--init")
 
       SSHKit::Backend::Abstract.any_instance.expects(:execute)
-        .with(:docker, :buildx, :build, "--push", "--platform", "linux/amd64", "--builder", "kamal-local-docker-container", "-t", "dhh/app:999", "-t", "dhh/app:latest", "--label", "service=\"app\"", "--file", "Dockerfile", ".")
+        .with(:docker, :buildx, :build, "--push", "--platform", "linux/amd64", "--builder", "kamal-local-docker-container", "-t", "dhh/app:999", "-t", "dhh/app:latest", "--label", "service=\"app\"", "--file", "Dockerfile", ".", env: {})
 
       SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
         .with(:git, "-C", anything, :"rev-parse", :HEAD)
@@ -140,7 +140,7 @@ class CliBuildTest < CliTestCase
         .returns("")
 
       SSHKit::Backend::Abstract.any_instance.expects(:execute)
-        .with(:docker, :buildx, :build, "--push", "--platform", "linux/amd64", "--builder", "kamal-local-docker-container", "-t", "dhh/app:999", "-t", "dhh/app:latest", "--label", "service=\"app\"", "--file", "Dockerfile", ".")
+        .with(:docker, :buildx, :build, "--push", "--platform", "linux/amd64", "--builder", "kamal-local-docker-container", "-t", "dhh/app:999", "-t", "dhh/app:latest", "--label", "service=\"app\"", "--file", "Dockerfile", ".", env: {})
 
       run_command("push").tap do |output|
         assert_match /WARN Missing compatible builder, so creating a new one first/, output

test/cli/cli_test_case.rb

@@ -29,19 +29,39 @@ class CliTestCase < ActiveSupport::TestCase
 
     def stub_setup
       SSHKit::Backend::Abstract.any_instance.stubs(:execute)
-        .with { |*args| args == [ :mkdir, "-p", ".kamal/apps/app" ] }
+        .with { |*args| args == [ :mkdir, "-p", ".kamal" ] }
       SSHKit::Backend::Abstract.any_instance.stubs(:execute)
-        .with { |arg1, arg2, arg3| arg1 == :mkdir && arg2 == "-p" && arg3 == ".kamal/lock-app" }
+        .with { |arg1, arg2, arg3| arg1 == :mkdir && arg2 == "-p" && arg3 == ".kamal/locks" }
       SSHKit::Backend::Abstract.any_instance.stubs(:execute)
-        .with { |arg1, arg2| arg1 == :mkdir && arg2 == ".kamal/lock-app" }
+        .with { |arg1, arg2| arg1 == :mkdir && arg2 == ".kamal/locks/app" }
       SSHKit::Backend::Abstract.any_instance.stubs(:execute)
-        .with { |arg1, arg2| arg1 == :rm && arg2 == ".kamal/lock-app/details" }
+        .with { |arg1, arg2| arg1 == :rm && arg2 == ".kamal/locks/app/details" }
       SSHKit::Backend::Abstract.any_instance.stubs(:execute)
         .with(:docker, :buildx, :inspect, "kamal-local-docker-container")
     end
 
     def assert_hook_ran(hook, output, version:, service_version:, hosts:, command:, subcommand: nil, runtime: false, secrets: false)
-      assert_match %r{usr/bin/env\s\.kamal/hooks/#{hook}}, output
+      whoami = `whoami`.chomp
+      performer = Kamal::Git.email.presence || whoami
+      service = service_version.split("@").first
+
+      assert_match "Running the #{hook} hook...\n", output
+
+      expected = %r{Running\s/usr/bin/env\s\.kamal/hooks/#{hook}\sas\s#{whoami}@localhost\n\s
+        DEBUG\s\[[0-9a-f]*\]\sCommand:\s\(\sexport\s
+        KAMAL_RECORDED_AT=\"\d\d\d\d-\d\d-\d\dT\d\d:\d\d:\d\dZ\"\s
+        KAMAL_PERFORMER=\"#{performer}\"\s
+        KAMAL_VERSION=\"#{version}\"\s
+        KAMAL_SERVICE_VERSION=\"#{service_version}\"\s
+        KAMAL_SERVICE=\"#{service}\"\s
+        KAMAL_HOSTS=\"#{hosts}\"\s
+        KAMAL_COMMAND=\"#{command}\"\s
+        #{"KAMAL_SUBCOMMAND=\\\"#{subcommand}\\\"\\s" if subcommand}
+        #{"KAMAL_RUNTIME=\\\"\\d+\\\"\\s" if runtime}
+        #{"DB_PASSWORD=\"secret\"\\s" if secrets}
+        ;\s/usr/bin/env\s\.kamal/hooks/#{hook} }x
+
+      assert_match expected, output
     end
 
     def with_argv(*argv)

test/cli/lock_test.rb

@@ -3,7 +3,7 @@ require_relative "cli_test_case"
 class CliLockTest < CliTestCase
   test "status" do
     run_command("status").tap do |output|
-      assert_match "Running /usr/bin/env stat .kamal/lock-app > /dev/null && cat .kamal/lock-app/details | base64 -d on 1.1.1.1", output
+      assert_match "Running /usr/bin/env stat .kamal/locks/app > /dev/null && cat .kamal/locks/app/details | base64 -d on 1.1.1.1", output
     end
   end
 

test/cli/main_test.rb

@@ -22,9 +22,9 @@ class CliMainTest < CliTestCase
     Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:server:bootstrap", [], 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:setup", [], invoke_options.merge(skip_local: true))
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:registry:login", [], invoke_options.merge(skip_local: true))
     Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:build:pull", [], invoke_options)
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:proxy:boot", [], 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)
@@ -35,7 +35,7 @@ class CliMainTest < CliTestCase
       assert_match /Acquiring the deploy lock/, output
       assert_match /Log into image registry/, output
       assert_match /Pull app image/, output
-      assert_match /Ensure kamal-proxy is running/, output
+      assert_match /Ensure Traefik is running/, output
       assert_match /Detect stale containers/, output
       assert_match /Prune old containers and images/, output
       assert_match /Releasing the deploy lock/, output
@@ -46,9 +46,9 @@ class CliMainTest < CliTestCase
     with_test_secrets("secrets" => "DB_PASSWORD=secret") do
       invoke_options = { "config_file" => "test/fixtures/deploy_simple.yml", "version" => "999", "skip_hooks" => false, "verbose" => true }
 
-      Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:registry:setup", [], invoke_options.merge(skip_local: false))
+      Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:registry:login", [], invoke_options.merge(skip_local: false))
       Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:build:deliver", [], invoke_options)
-      Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:proxy:boot", [], 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)
@@ -61,7 +61,7 @@ class CliMainTest < CliTestCase
         assert_match /Log into image registry/, output
         assert_match /Build and push app image/, output
         assert_hook_ran "pre-deploy", output, **hook_variables, secrets: true
-        assert_match /Ensure kamal-proxy is running/, output
+        assert_match /Ensure Traefik is running/, output
         assert_match /Detect stale containers/, output
         assert_match /Prune old containers and images/, output
         assert_hook_ran "post-deploy", output, **hook_variables, runtime: true, secrets: true
@@ -72,9 +72,9 @@ class CliMainTest < CliTestCase
   test "deploy 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:registry:setup", [], invoke_options.merge(skip_local: true))
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:registry:login", [], invoke_options.merge(skip_local: true))
     Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:build:pull", [], invoke_options)
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:proxy:boot", [], 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)
@@ -83,7 +83,7 @@ class CliMainTest < CliTestCase
       assert_match /Acquiring the deploy lock/, output
       assert_match /Log into image registry/, output
       assert_match /Pull app image/, output
-      assert_match /Ensure kamal-proxy is running/, output
+      assert_match /Ensure Traefik is running/, output
       assert_match /Detect stale containers/, output
       assert_match /Prune old containers and images/, output
       assert_match /Releasing the deploy lock/, output
@@ -97,14 +97,17 @@ class CliMainTest < CliTestCase
     Dir.stubs(:chdir)
 
     SSHKit::Backend::Abstract.any_instance.stubs(:execute)
-      .with { |*args| args == [ :mkdir, "-p", ".kamal/apps/app" ] }
+      .with { |*args| args == [ :mkdir, "-p", ".kamal" ] }
 
     SSHKit::Backend::Abstract.any_instance.stubs(:execute)
-      .with { |*arg| arg[0..1] == [ :mkdir, ".kamal/lock-app" ] }
-      .raises(RuntimeError, "mkdir: cannot create directory ‘kamal/lock-app’: File exists")
+      .with { |*args| args == [ :mkdir, "-p", ".kamal/locks" ] }
+
+    SSHKit::Backend::Abstract.any_instance.stubs(:execute)
+      .with { |*arg| arg[0..1] == [ :mkdir, ".kamal/locks/app" ] }
+      .raises(RuntimeError, "mkdir: cannot create directory ‘kamal/locks/app’: File exists")
 
     SSHKit::Backend::Abstract.any_instance.expects(:capture_with_debug)
-      .with(:stat, ".kamal/lock-app", ">", "/dev/null", "&&", :cat, ".kamal/lock-app/details", "|", :base64, "-d")
+      .with(:stat, ".kamal/locks/app", ">", "/dev/null", "&&", :cat, ".kamal/locks/app/details", "|", :base64, "-d")
 
     SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
       .with(:git, "-C", anything, :"rev-parse", :HEAD)
@@ -131,10 +134,13 @@ class CliMainTest < CliTestCase
     Dir.stubs(:chdir)
 
     SSHKit::Backend::Abstract.any_instance.stubs(:execute)
-      .with { |*args| args == [ :mkdir, "-p", ".kamal/apps/app" ] }
+      .with { |*args| args == [ :mkdir, "-p", ".kamal" ] }
 
     SSHKit::Backend::Abstract.any_instance.stubs(:execute)
-      .with { |*arg| arg[0..1] == [ :mkdir, ".kamal/lock-app" ] }
+      .with { |*args| args == [ :mkdir, "-p", ".kamal/locks" ] }
+
+    SSHKit::Backend::Abstract.any_instance.stubs(:execute)
+      .with { |*arg| arg[0..1] == [ :mkdir, ".kamal/locks/app" ] }
       .raises(SocketError, "getaddrinfo: nodename nor servname provided, or not known")
 
     SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
@@ -159,7 +165,7 @@ class CliMainTest < CliTestCase
     invoke_options = { "config_file" => "test/fixtures/deploy_simple.yml", "version" => "999", "skip_hooks" => false, :skip_local => false }
 
     Kamal::Cli::Main.any_instance.expects(:invoke)
-      .with("kamal:cli:registry:setup", [], invoke_options.merge(skip_local: false))
+      .with("kamal:cli:registry:login", [], invoke_options.merge(skip_local: false))
       .raises(RuntimeError)
 
     assert_not KAMAL.holding_lock?
@@ -172,9 +178,9 @@ class CliMainTest < CliTestCase
   test "deploy with skipped hooks" do
     invoke_options = { "config_file" => "test/fixtures/deploy_simple.yml", "version" => "999", "skip_hooks" => true }
 
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:registry:setup", [], invoke_options.merge(skip_local: false))
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:registry:login", [], invoke_options.merge(skip_local: false))
     Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:build:deliver", [], invoke_options)
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:proxy:boot", [], 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)
@@ -184,12 +190,27 @@ class CliMainTest < CliTestCase
     end
   end
 
+  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.merge(skip_local: false))
+    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:setup", [], invoke_options.merge(skip_local: false))
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:registry:login", [], invoke_options.merge(skip_local: false))
     Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:build:deliver", [], invoke_options)
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:proxy:boot", [], 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)
@@ -252,11 +273,18 @@ class CliMainTest < CliTestCase
       SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
         .with(:sh, "-c", "'docker ps --latest --format '\\''{{.Names}}'\\'' --filter label=service=app --filter label=role=#{role} --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --format '\\''{{.Names}}'\\'' --filter label=service=app --filter label=role=#{role} --filter status=running --filter status=restarting'", "|", :head, "-1", "|", "while read line; do echo ${line#app-#{role}-}; done", raise_on_non_zero_exit: false)
         .returns("version-to-rollback\n").at_least_once
+      SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
+        .with(:docker, :container, :ls, "--all", "--filter", "name=^app-#{role}-123$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
+        .returns("running").at_least_once # health check
     end
 
     SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
-      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-workers-123$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
-      .returns("running").at_least_once # health check
+      .with(:docker, :inspect, "-f '{{ range .Mounts }}{{printf \"%s %s\\n\" .Source .Destination}}{{ end }}'", "app-web-version-to-rollback", "|", :awk, "'$2 == \"/tmp/kamal-cord\" {print $1}'", raise_on_non_zero_exit: false)
+      .returns("corddirectory").at_least_once # health check
+
+    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
+      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-version-to-rollback$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
+      .returns("unhealthy").at_least_once # health check
 
     Kamal::Commands::Hook.any_instance.stubs(:hook_exists?).returns(true)
     hook_variables = { version: 123, service_version: "app@123", hosts: "1.1.1.1,1.1.1.2,1.1.1.3,1.1.1.4", command: "rollback" }
@@ -273,15 +301,17 @@ class CliMainTest < CliTestCase
   test "rollback without old version" do
     Kamal::Cli::Main.any_instance.stubs(:container_available?).returns(true)
 
+    Kamal::Cli::Healthcheck::Poller.stubs(:sleep)
+
     SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
       .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-123$", "--quiet", raise_on_non_zero_exit: false)
       .returns("").at_least_once
-    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
-      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-123$", "--quiet")
-      .returns("123").at_least_once
     SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
       .with(:sh, "-c", "'docker ps --latest --format '\\''{{.Names}}'\\'' --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --format '\\''{{.Names}}'\\'' --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting'", "|", :head, "-1", "|", "while read line; do echo ${line#app-web-}; done", raise_on_non_zero_exit: false)
       .returns("").at_least_once
+    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
+      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-123$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
+      .returns("running").at_least_once # health check
 
     run_command("rollback", "123").tap do |output|
       assert_match "docker run --detach --restart unless-stopped --name app-web-123", output
@@ -289,18 +319,8 @@ class CliMainTest < CliTestCase
     end
   end
 
-  test "remove" do
-    options = { "config_file" => "test/fixtures/deploy_simple.yml", "skip_hooks" => false, "confirmed" => true }
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:proxy:remove", [], options)
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:app:remove", [], options)
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:accessory:remove", [ "all" ], options)
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:registry:remove", [], options.merge(skip_local: true))
-
-    run_command("remove", "-y")
-  end
-
   test "details" do
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:proxy:details")
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:traefik:details")
     Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:app:details")
     Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:accessory:details", [ "all" ])
 
@@ -414,14 +434,13 @@ class CliMainTest < CliTestCase
 
   test "remove with confirmation" do
     run_command("remove", "-y", config_file: "deploy_with_accessories").tap do |output|
-      assert_match /docker container stop kamal-proxy/, output
-      assert_match /docker container prune --force --filter label=org.opencontainers.image.title=kamal-proxy/, output
-      assert_match /docker image prune --all --force --filter label=org.opencontainers.image.title=kamal-proxy/, 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 image prune --all --force --filter label=org.opencontainers.image.title=Traefik/, output
 
       assert_match /docker ps --quiet --filter label=service=app | xargs docker stop/, output
       assert_match /docker container prune --force --filter label=service=app/, output
       assert_match /docker image prune --all --force --filter label=service=app/, output
-      assert_match "/usr/bin/env rm -r .kamal/apps/app", output
 
       assert_match /docker container stop app-mysql/, output
       assert_match /docker container prune --force --filter label=service=app-mysql/, output
@@ -461,7 +480,7 @@ class CliMainTest < CliTestCase
   end
 
   test "run an alias for details" do
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:proxy:details")
+    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:traefik:details")
     Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:app:details")
     Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:accessory:details", [ "all" ])
 
@@ -496,34 +515,6 @@ class CliMainTest < CliTestCase
     end
   end
 
-  test "upgrade" do
-    invoke_options = { "config_file" => "test/fixtures/deploy_with_accessories.yml", "skip_hooks" => false, "confirmed" => true, "rolling" => false }
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:proxy:upgrade", [], invoke_options)
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:accessory:upgrade", [ "all" ], invoke_options)
-
-    run_command("upgrade", "-y", config_file: "deploy_with_accessories").tap do |output|
-      assert_match "Upgrading all hosts...", output
-      assert_match "Upgraded all hosts", output
-    end
-  end
-
-  test "upgrade rolling" do
-    invoke_options = { "config_file" => "test/fixtures/deploy_with_accessories.yml", "skip_hooks" => false, "confirmed" => true, "rolling" => false }
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:proxy:upgrade", [], invoke_options).times(4)
-    Kamal::Cli::Main.any_instance.expects(:invoke).with("kamal:cli:accessory:upgrade", [ "all" ], invoke_options).times(3)
-
-    run_command("upgrade", "--rolling", "-y", config_file: "deploy_with_accessories").tap do |output|
-      assert_match "Upgrading 1.1.1.1...", output
-      assert_match "Upgraded 1.1.1.1", output
-      assert_match "Upgrading 1.1.1.2...", output
-      assert_match "Upgraded 1.1.1.2", output
-      assert_match "Upgrading 1.1.1.3...", output
-      assert_match "Upgraded 1.1.1.3", output
-      assert_match "Upgrading 1.1.1.4...", output
-      assert_match "Upgraded 1.1.1.4", output
-    end
-  end
-
   private
     def run_command(*command, config_file: "deploy_simple")
       with_argv([ *command, "-c", "test/fixtures/#{config_file}.yml" ]) do

test/cli/proxy_test.rb

@@ -4,44 +4,10 @@ class CliProxyTest < CliTestCase
   test "boot" do
     run_command("boot").tap do |output|
       assert_match "docker login", output
-      assert_match "docker run --name kamal-proxy --network kamal --detach --restart unless-stopped --publish 80:80 --publish 443:443 --volume kamal-proxy-config:/home/kamal-proxy/.config/kamal-proxy --log-opt max-size=\"10m\" #{KAMAL.config.proxy_image}", output
+      assert_match "docker run --name kamal-proxy --network kamal --detach --restart unless-stopped --publish 80:80 --publish 443:443 --volume /var/run/docker.sock:/var/run/docker.sock --volume $(pwd)/.kamal/proxy/config:/root/.config/kamal-proxy --log-opt max-size=\"10m\" #{Kamal::Configuration::Proxy::DEFAULT_IMAGE}", output
     end
   end
 
-  test "boot old version" do
-    Thread.report_on_exception = false
-    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
-      .with(:docker, :inspect, "kamal-proxy", "--format '{{.Config.Image}}'", "|", :cut, "-d:", "-f2")
-      .returns("v0.0.1")
-      .at_least_once
-
-    exception = assert_raises do
-      run_command("boot").tap do |output|
-        assert_match "docker login", output
-        assert_match "docker run --name kamal-proxy --network kamal --detach --restart unless-stopped --publish 80:80 --publish 443:443 --volume kamal-proxy-config:/home/kamal-proxy/.config/kamal-proxy --log-opt max-size=\"10m\" #{KAMAL.config.proxy_image}", output
-      end
-    end
-
-    assert_includes exception.message, "kamal-proxy version v0.0.1 is too old, please reboot to update to at least #{Kamal::Configuration::PROXY_MINIMUM_VERSION}"
-  ensure
-    Thread.report_on_exception = false
-  end
-
-  test "boot correct version" do
-    Thread.report_on_exception = false
-    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
-      .with(:docker, :inspect, "kamal-proxy", "--format '{{.Config.Image}}'", "|", :cut, "-d:", "-f2")
-      .returns(Kamal::Configuration::PROXY_MINIMUM_VERSION)
-      .at_least_once
-
-    run_command("boot").tap do |output|
-      assert_match "docker login", output
-      assert_match "docker container start kamal-proxy || docker run --name kamal-proxy --network kamal --detach --restart unless-stopped --publish 80:80 --publish 443:443 --volume kamal-proxy-config:/home/kamal-proxy/.config/kamal-proxy --log-opt max-size=\"10m\" #{KAMAL.config.proxy_image}", output
-    end
-  ensure
-    Thread.report_on_exception = false
-  end
-
   test "reboot" do
     SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
       .with(:docker, :container, :ls, "--all", "--filter", "name=^app-web-123$", "--quiet")
@@ -55,16 +21,17 @@ class CliProxyTest < CliTestCase
 
     run_command("reboot", "-y").tap do |output|
       assert_match "docker container stop kamal-proxy on 1.1.1.1", output
-      assert_match "Running docker container stop traefik ; docker container prune --force --filter label=org.opencontainers.image.title=Traefik && docker image prune --all --force --filter label=org.opencontainers.image.title=Traefik on 1.1.1.1", output
+      assert_match "docker container stop traefik on 1.1.1.1", output
       assert_match "docker container prune --force --filter label=org.opencontainers.image.title=kamal-proxy on 1.1.1.1", output
-      assert_match "docker run --name kamal-proxy --network kamal --detach --restart unless-stopped --publish 80:80 --publish 443:443 --volume kamal-proxy-config:/home/kamal-proxy/.config/kamal-proxy --log-opt max-size=\"10m\" #{KAMAL.config.proxy_image} on 1.1.1.1", output
-      assert_match "docker exec kamal-proxy kamal-proxy deploy app-web --target \"abcdefabcdef:80\" --deploy-timeout \"6s\" --drain-timeout \"30s\" --buffer-requests --buffer-responses --log-request-header \"Cache-Control\" --log-request-header \"Last-Modified\" --log-request-header \"User-Agent\" on 1.1.1.1", output
+      assert_match "docker container prune --force --filter label=org.opencontainers.image.title=Traefik on 1.1.1.1", output
+      assert_match "docker run --name kamal-proxy --network kamal --detach --restart unless-stopped --publish 80:80 --publish 443:443 --volume /var/run/docker.sock:/var/run/docker.sock --volume $(pwd)/.kamal/proxy/config:/root/.config/kamal-proxy --log-opt max-size=\"10m\" #{Kamal::Configuration::Proxy::DEFAULT_IMAGE} on 1.1.1.1", output
+      assert_match "docker exec kamal-proxy kamal-proxy deploy app-web --target \"abcdefabcdef:80\" --deploy-timeout \"6s\" --buffer-requests --buffer-responses --log-request-header \"Cache-Control\" --log-request-header \"Last-Modified\" on 1.1.1.1", output
 
       assert_match "docker container stop kamal-proxy on 1.1.1.2", output
-      assert_match "Running docker container stop traefik ; docker container prune --force --filter label=org.opencontainers.image.title=Traefik && docker image prune --all --force --filter label=org.opencontainers.image.title=Traefik on 1.1.1.2", output
+      assert_match "docker container stop traefik on 1.1.1.2", output
       assert_match "docker container prune --force --filter label=org.opencontainers.image.title=kamal-proxy on 1.1.1.2", output
-      assert_match "docker run --name kamal-proxy --network kamal --detach --restart unless-stopped --publish 80:80 --publish 443:443 --volume kamal-proxy-config:/home/kamal-proxy/.config/kamal-proxy --log-opt max-size=\"10m\" #{KAMAL.config.proxy_image} on 1.1.1.2", output
-      assert_match "docker exec kamal-proxy kamal-proxy deploy app-web --target \"abcdefabcdef:80\" --deploy-timeout \"6s\" --drain-timeout \"30s\" --buffer-requests --buffer-responses --log-request-header \"Cache-Control\" --log-request-header \"Last-Modified\" --log-request-header \"User-Agent\" on 1.1.1.2", output
+      assert_match "docker container prune --force --filter label=org.opencontainers.image.title=Traefik on 1.1.1.2", 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\" traefik:v2.10 --providers.docker --log.level=\"DEBUG\" on 1.1.1.2", output
     end
   end
 
@@ -111,11 +78,11 @@ class CliProxyTest < CliTestCase
 
   test "logs" do
     SSHKit::Backend::Abstract.any_instance.stubs(:capture)
-      .with(:docker, :logs, "kamal-proxy", "--tail 100", "--timestamps", "2>&1")
+      .with(:docker, :logs, "kamal-proxy", " --tail 100", "--timestamps", "2>&1")
       .returns("Log entry")
 
     SSHKit::Backend::Abstract.any_instance.stubs(:capture)
-      .with(:docker, :logs, "proxy", "--tail 100", "--timestamps", "2>&1")
+      .with(:docker, :logs, "traefik", " --tail 100", "--timestamps", "2>&1")
       .returns("Log entry")
 
     run_command("logs").tap do |output|
@@ -132,35 +99,11 @@ class CliProxyTest < CliTestCase
   end
 
   test "remove" do
-    run_command("remove").tap do |output|
-      assert_match "/usr/bin/env ls .kamal/apps | wc -l", output
-      assert_match "docker container prune --force --filter label=org.opencontainers.image.title=kamal-proxy", output
-      assert_match "docker image prune --all --force --filter label=org.opencontainers.image.title=kamal-proxy", output
-    end
-  end
+    Kamal::Cli::Proxy.any_instance.expects(:stop)
+    Kamal::Cli::Proxy.any_instance.expects(:remove_container)
+    Kamal::Cli::Proxy.any_instance.expects(:remove_image)
 
-  test "remove with other apps" do
-    Thread.report_on_exception = false
-
-    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info).with(:ls, ".kamal/apps", "|", :wc, "-l").returns("1\n").twice
-
-    run_command("remove").tap do |output|
-      assert_match "Not removing the proxy, as other apps are installed, ignore this check with kamal proxy remove --force", output
-    end
-  ensure
-    Thread.report_on_exception = true
-  end
-
-  test "force remove with other apps" do
-    Thread.report_on_exception = false
-
-    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info).with(:ls, ".kamal/apps", "|", :wc, "-l").returns("1\n").twice
-
-    run_command("remove").tap do |output|
-      assert_match "Not removing the proxy, as other apps are installed, ignore this check with kamal proxy remove --force", output
-    end
-  ensure
-    Thread.report_on_exception = true
+    run_command("remove")
   end
 
   test "remove_container" do
@@ -175,69 +118,24 @@ class CliProxyTest < CliTestCase
     end
   end
 
-  test "upgrade" do
-    Object.any_instance.stubs(:sleep)
-
-    SSHKit::Backend::Abstract.any_instance.stubs(:capture_with_info).returns("12345678")
-
-    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
-      .with(:docker, :inspect, "kamal-proxy", "--format '{{.Config.Image}}'", "|", :cut, "-d:", "-f2")
-      .returns(Kamal::Configuration::PROXY_MINIMUM_VERSION)
-
-    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
-      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-workers-latest$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
-      .returns("running").at_least_once # workers health check
-
-    run_command("upgrade", "-y").tap do |output|
-      assert_match "Upgrading proxy on 1.1.1.1,1.1.1.2,1.1.1.3,1.1.1.4...", output
-      assert_match "docker login -u [REDACTED] -p [REDACTED]", output
-      assert_match "docker container stop traefik ; docker container prune --force --filter label=org.opencontainers.image.title=Traefik && docker image prune --all --force --filter label=org.opencontainers.image.title=Traefik", output
-      assert_match "docker container stop kamal-proxy", output
-      assert_match "docker container prune --force --filter label=org.opencontainers.image.title=kamal-proxy", output
-      assert_match "docker image prune --all --force --filter label=org.opencontainers.image.title=kamal-proxy", output
-      assert_match "/usr/bin/env mkdir -p .kamal", output
-      assert_match "docker network create kamal", output
-      assert_match "docker login -u [REDACTED] -p [REDACTED]", output
-      assert_match "docker container start kamal-proxy || docker run --name kamal-proxy --network kamal --detach --restart unless-stopped --publish 80:80 --publish 443:443 --volume kamal-proxy-config:/home/kamal-proxy/.config/kamal-proxy --log-opt max-size=\"10m\" basecamp/kamal-proxy:#{Kamal::Configuration::PROXY_MINIMUM_VERSION}", output
-      assert_match "/usr/bin/env mkdir -p .kamal", output
-      assert_match %r{docker rename app-web-latest app-web-latest_replaced_.*}, output
-      assert_match "/usr/bin/env mkdir -p .kamal/apps/app/env/roles", output
-      assert_match %r{/usr/bin/env .* .kamal/apps/app/env/roles/web.env}, output
-      assert_match %r{docker run --detach --restart unless-stopped --name app-web-latest --network kamal --hostname 1.1.1.1-.* -e KAMAL_CONTAINER_NAME="app-web-latest" -e KAMAL_VERSION="latest" --env-file .kamal/apps/app/env/roles/web.env --log-opt max-size="10m" --label service="app" --label role="web" --label destination dhh/app:latest}, output
-      assert_match "docker exec kamal-proxy kamal-proxy deploy app-web --target \"12345678:80\" --deploy-timeout \"6s\" --drain-timeout \"30s\" --buffer-requests --buffer-responses --log-request-header \"Cache-Control\" --log-request-header \"Last-Modified\" --log-request-header \"User-Agent\"", output
-      assert_match "docker container ls --all --filter name=^app-web-12345678$ --quiet | xargs docker stop", output
-      assert_match "docker tag dhh/app:latest dhh/app:latest", output
-      assert_match "/usr/bin/env mkdir -p .kamal", output
-      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 $container_id; done", output
-      assert_match "docker image prune --force --filter label=service=app", output
-      assert_match "Upgraded proxy on 1.1.1.1,1.1.1.2,1.1.1.3,1.1.1.4", output
-    end
-  end
-
-  test "upgrade rolling" do
-    Object.any_instance.stubs(:sleep)
-
-    SSHKit::Backend::Abstract.any_instance.stubs(:capture_with_info).returns("12345678")
-
-    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
-      .with(:docker, :inspect, "kamal-proxy", "--format '{{.Config.Image}}'", "|", :cut, "-d:", "-f2")
-      .returns(Kamal::Configuration::PROXY_MINIMUM_VERSION)
-
-    SSHKit::Backend::Abstract.any_instance.expects(:capture_with_info)
-      .with(:docker, :container, :ls, "--all", "--filter", "name=^app-workers-latest$", "--quiet", "|", :xargs, :docker, :inspect, "--format", "'{{if .State.Health}}{{.State.Health.Status}}{{else}}{{.State.Status}}{{end}}'")
-      .returns("running").at_least_once # workers health check
-
-    run_command("upgrade", "--rolling", "-y",).tap do |output|
-      %w[1.1.1.1 1.1.1.2 1.1.1.3 1.1.1.4].each do |host|
-        assert_match "Upgrading proxy on #{host}...", output
-        assert_match "docker container stop traefik ; docker container prune --force --filter label=org.opencontainers.image.title=Traefik && docker image prune --all --force --filter label=org.opencontainers.image.title=Traefik on #{host}", output
-        assert_match "Upgraded proxy on #{host}", output
-      end
-    end
+  test "commands disallowed when proxy is disabled" do
+    assert_raises_when_disabled "boot"
+    assert_raises_when_disabled "reboot"
+    assert_raises_when_disabled "start"
+    assert_raises_when_disabled "stop"
+    assert_raises_when_disabled "details"
+    assert_raises_when_disabled "logs"
+    assert_raises_when_disabled "remove"
   end
 
   private
     def run_command(*command, fixture: :with_proxy)
       stdouted { Kamal::Cli::Proxy.start([ *command, "-c", "test/fixtures/deploy_#{fixture}.yml" ]) }
     end
+
+    def assert_raises_when_disabled(command)
+      assert_raises "kamal proxy commands are disabled unless experimental proxy support is enabled. Use `kamal traefik` commands instead." do
+        run_command(command, fixture: :with_accessories)
+      end
+    end
 end

test/cli/prune_test.rb

@@ -18,10 +18,12 @@ class CliPruneTest < CliTestCase
   test "containers" do
     run_command("containers").tap do |output|
       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

test/cli/registry_test.rb

@@ -1,62 +1,50 @@
 require_relative "cli_test_case"
 
 class CliRegistryTest < CliTestCase
-  test "setup" do
-    run_command("setup").tap do |output|
+  test "login" do
+    run_command("login").tap do |output|
       assert_match /docker login -u \[REDACTED\] -p \[REDACTED\] as .*@localhost/, output
       assert_match /docker login -u \[REDACTED\] -p \[REDACTED\] on 1.1.1.\d/, output
     end
   end
 
-  test "setup skip local" do
-    run_command("setup", "-L").tap do |output|
+  test "login skip local" do
+    run_command("login", "-L").tap do |output|
       assert_no_match /docker login -u \[REDACTED\] -p \[REDACTED\] as .*@localhost/, output
       assert_match /docker login -u \[REDACTED\] -p \[REDACTED\] on 1.1.1.\d/, output
     end
   end
 
-  test "setup skip remote" do
-    run_command("setup", "-R").tap do |output|
+  test "login skip remote" do
+    run_command("login", "-R").tap do |output|
       assert_match /docker login -u \[REDACTED\] -p \[REDACTED\] as .*@localhost/, output
       assert_no_match /docker login -u \[REDACTED\] -p \[REDACTED\] on 1.1.1.\d/, output
     end
   end
 
-  test "remove" do
-    run_command("remove").tap do |output|
+  test "logout" do
+    run_command("logout").tap do |output|
       assert_match /docker logout as .*@localhost/, output
       assert_match /docker logout on 1.1.1.\d/, output
     end
   end
 
-  test "remove skip local" do
-    run_command("remove", "-L").tap do |output|
+  test "logout skip local" do
+    run_command("logout", "-L").tap do |output|
       assert_no_match /docker logout as .*@localhost/, output
       assert_match /docker logout on 1.1.1.\d/, output
     end
   end
 
-  test "remove skip remote" do
-    run_command("remove", "-R").tap do |output|
+  test "logout skip remote" do
+    run_command("logout", "-R").tap do |output|
       assert_match /docker logout as .*@localhost/, output
       assert_no_match /docker logout on 1.1.1.\d/, output
     end
   end
 
-  test "setup local registry" do
-    run_command("setup", fixture: :with_local_registry).tap do |output|
-      assert_match /docker start kamal-docker-registry || docker run --detach -p 127.0.0.1:5000:5000 --name kamal-docker-registry registry:2 as .*@localhost/, output
-    end
-  end
-
-  test "remove local registry" do
-    run_command("remove", fixture: :with_local_registry).tap do |output|
-      assert_match /docker stop kamal-docker-registry && docker rm kamal-docker-registry as .*@localhost/, output
-    end
-  end
-
   private
-    def run_command(*command, fixture: :with_accessories)
-      stdouted { Kamal::Cli::Registry.start([ *command, "-c", "test/fixtures/deploy_#{fixture}.yml" ]) }
+    def run_command(*command)
+      stdouted { Kamal::Cli::Registry.start([ *command, "-c", "test/fixtures/deploy_with_accessories.yml" ]) }
     end
 end

test/cli/traefik_test.rb

@@ -0,0 +1,110 @@
+require_relative "cli_test_case"
+
+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\" --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::Configuration::Traefik::DEFAULT_IMAGE} --providers.docker --log.level=\"DEBUG\"", output
+    end
+  end
+
+  test "reboot" do
+    Kamal::Commands::Registry.any_instance.expects(:login).twice
+
+    run_command("reboot", "-y").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\" --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::Configuration::Traefik::DEFAULT_IMAGE} --providers.docker --log.level=\"DEBUG\"", output
+    end
+  end
+
+  test "reboot --rolling" do
+    Object.any_instance.stubs(:sleep)
+
+    run_command("reboot", "--rolling", "-y").tap do |output|
+      assert_match "Running docker container prune --force --filter label=org.opencontainers.image.title=Traefik on 1.1.1.1", output
+    end
+  end
+
+  test "start" do
+    run_command("start").tap do |output|
+      assert_match "docker container start traefik", output
+    end
+  end
+
+  test "stop" do
+    run_command("stop").tap do |output|
+      assert_match "docker container stop traefik", output
+    end
+  end
+
+  test "restart" do
+    Kamal::Cli::Traefik.any_instance.expects(:stop)
+    Kamal::Cli::Traefik.any_instance.expects(:start)
+
+    run_command("restart")
+  end
+
+  test "details" do
+    run_command("details").tap do |output|
+      assert_match "docker ps --filter name=^traefik$", output
+    end
+  end
+
+  test "logs" do
+    SSHKit::Backend::Abstract.any_instance.stubs(:capture)
+      .with(:docker, :logs, "traefik", " --tail 100", "--timestamps", "2>&1")
+      .returns("Log entry")
+
+    run_command("logs").tap do |output|
+      assert_match "Traefik Host: 1.1.1.1", output
+      assert_match "Log entry", output
+    end
+  end
+
+  test "logs with follow" do
+    SSHKit::Backend::Abstract.any_instance.stubs(:exec)
+      .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 "logs with follow and grep" do
+    SSHKit::Backend::Abstract.any_instance.stubs(:exec)
+      .with("ssh -t root@1.1.1.1 -p 22 'docker logs traefik --timestamps --tail 10 --follow 2>&1 | grep \"hey\"'")
+
+    assert_match "docker logs traefik --timestamps --tail 10 --follow 2>&1 | grep \"hey\"", run_command("logs", "--follow", "--grep", "hey")
+  end
+
+  test "logs with follow, grep, and grep options" do
+    SSHKit::Backend::Abstract.any_instance.stubs(:exec)
+      .with("ssh -t root@1.1.1.1 -p 22 'docker logs traefik --timestamps --tail 10 --follow 2>&1 | grep \"hey\" -C 2'")
+
+    assert_match "docker logs traefik --timestamps --tail 10 --follow 2>&1 | grep \"hey\" -C 2", run_command("logs", "--follow", "--grep", "hey", "--grep-options", "-C 2")
+  end
+
+  test "remove" do
+    Kamal::Cli::Traefik.any_instance.expects(:stop)
+    Kamal::Cli::Traefik.any_instance.expects(:remove_container)
+    Kamal::Cli::Traefik.any_instance.expects(:remove_image)
+
+    run_command("remove")
+  end
+
+  test "remove_container" do
+    run_command("remove_container").tap do |output|
+      assert_match "docker container prune --force --filter label=org.opencontainers.image.title=Traefik", output
+    end
+  end
+
+  test "remove_image" do
+    run_command("remove_image").tap do |output|
+      assert_match "docker image prune --all --force --filter label=org.opencontainers.image.title=Traefik", output
+    end
+  end
+
+  private
+    def run_command(*command)
+      stdouted { Kamal::Cli::Traefik.start([ *command, "-c", "test/fixtures/deploy_with_accessories.yml" ]) }
+    end
+end

test/commander_test.rb

@@ -136,18 +136,18 @@ class CommanderTest < ActiveSupport::TestCase
     assert_equal [ "1.1.1.3", "1.1.1.4", "1.1.1.1", "1.1.1.2" ], @kamal.hosts
   end
 
-  test "proxy hosts should observe filtered roles" do
-    configure_with(:deploy_with_multiple_proxy_roles)
+  test "traefik hosts should observe filtered roles" do
+    configure_with(:deploy_with_multiple_traefik_roles)
 
     @kamal.specific_roles = [ "web_tokyo" ]
-    assert_equal [ "1.1.1.3", "1.1.1.4" ], @kamal.proxy_hosts
+    assert_equal [ "1.1.1.3", "1.1.1.4" ], @kamal.traefik_hosts
   end
 
-  test "proxy hosts should observe filtered hosts" do
-    configure_with(:deploy_with_multiple_proxy_roles)
+  test "traefik hosts should observe filtered hosts" do
+    configure_with(:deploy_with_multiple_traefik_roles)
 
     @kamal.specific_hosts = [ "1.1.1.2" ]
-    assert_equal [ "1.1.1.2" ], @kamal.proxy_hosts
+    assert_equal [ "1.1.1.2" ], @kamal.traefik_hosts
   end
 
   private

test/commands/accessory_test.rb

@@ -51,15 +51,15 @@ class CommandsAccessoryTest < ActiveSupport::TestCase
 
   test "run" do
     assert_equal \
-      "docker run --name app-mysql --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --publish 3306:3306 --env MYSQL_ROOT_HOST=\"%\" --env-file .kamal/apps/app/env/accessories/mysql.env --label service=\"app-mysql\" private.registry/mysql:8.0",
+      "docker run --name app-mysql --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --publish 3306:3306 --env MYSQL_ROOT_HOST=\"%\" --env-file .kamal/env/accessories/app-mysql.env --label service=\"app-mysql\" private.registry/mysql:8.0",
       new_command(:mysql).run.join(" ")
 
     assert_equal \
-      "docker run --name app-redis --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --publish 6379:6379 --env SOMETHING=\"else\" --env-file .kamal/apps/app/env/accessories/redis.env --volume /var/lib/redis:/data --label service=\"app-redis\" --label cache=\"true\" redis:latest",
+      "docker run --name app-redis --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --publish 6379:6379 --env SOMETHING=\"else\" --env-file .kamal/env/accessories/app-redis.env --volume /var/lib/redis:/data --label service=\"app-redis\" --label cache=\"true\" redis:latest",
       new_command(:redis).run.join(" ")
 
     assert_equal \
-      "docker run --name custom-busybox --detach --restart unless-stopped --network kamal --log-opt max-size=\"10m\" --env-file .kamal/apps/app/env/accessories/busybox.env --label service=\"custom-busybox\" busybox:latest",
+      "docker run --name custom-busybox --detach --restart unless-stopped --network kamal --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
 
@@ -67,7 +67,7 @@ class CommandsAccessoryTest < ActiveSupport::TestCase
     @config[:logging] = { "driver" => "local", "options" => { "max-size" => "100m", "max-file" => "3" } }
 
     assert_equal \
-      "docker run --name custom-busybox --detach --restart unless-stopped --network kamal --log-driver \"local\" --log-opt max-size=\"100m\" --log-opt max-file=\"3\" --env-file .kamal/apps/app/env/accessories/busybox.env --label service=\"custom-busybox\" busybox:latest",
+      "docker run --name custom-busybox --detach --restart unless-stopped --network kamal --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
 
@@ -92,7 +92,7 @@ class CommandsAccessoryTest < ActiveSupport::TestCase
 
   test "execute in new container" do
     assert_equal \
-      "docker run --rm --network kamal --env MYSQL_ROOT_HOST=\"%\" --env-file .kamal/apps/app/env/accessories/mysql.env private.registry/mysql:8.0 mysql -u root",
+      "docker run --rm --network kamal --env MYSQL_ROOT_HOST=\"%\" --env-file .kamal/env/accessories/app-mysql.env private.registry/mysql:8.0 mysql -u root",
       new_command(:mysql).execute_in_new_container("mysql", "-u", "root").join(" ")
   end
 
@@ -104,7 +104,7 @@ class CommandsAccessoryTest < ActiveSupport::TestCase
 
   test "execute in new container over ssh" do
     new_command(:mysql).stub(:run_over_ssh, ->(cmd) { cmd.join(" ") }) do
-      assert_match %r{docker run -it --rm --network kamal --env MYSQL_ROOT_HOST=\"%\" --env-file .kamal/apps/app/env/accessories/mysql.env private.registry/mysql:8.0 mysql -u root},
+      assert_match %r{docker run -it --rm --network kamal --env MYSQL_ROOT_HOST=\"%\" --env-file .kamal/env/accessories/app-mysql.env private.registry/mysql:8.0 mysql -u root},
         new_command(:mysql).execute_in_new_container_over_ssh("mysql", "-u", "root")
     end
   end
@@ -130,20 +130,12 @@ class CommandsAccessoryTest < ActiveSupport::TestCase
     assert_equal \
       "docker logs app-mysql  --since 5m  --tail 100 --timestamps 2>&1 | grep 'thing' -C 2",
       new_command(:mysql).logs(since: "5m", lines: 100, grep: "thing", grep_options: "-C 2").join(" ")
-
-    assert_equal \
-      "docker logs app-mysql  --since 5m  --tail 100 2>&1 | grep 'thing' -C 2",
-      new_command(:mysql).logs(timestamps: false, since: "5m", lines: 100, grep: "thing", grep_options: "-C 2").join(" ")
   end
 
   test "follow logs" do
     assert_equal \
       "ssh -t root@1.1.1.5 -p 22 'docker logs app-mysql --timestamps --tail 10 --follow 2>&1'",
       new_command(:mysql).follow_logs
-
-    assert_equal \
-      "ssh -t root@1.1.1.5 -p 22 'docker logs app-mysql --tail 10 --follow 2>&1'",
-      new_command(:mysql).follow_logs(timestamps: false)
   end
 
   test "remove container" do

test/commands/app_test.rb

@@ -3,8 +3,9 @@ require "test_helper"
 class CommandsAppTest < ActiveSupport::TestCase
   setup do
     setup_test_secrets("secrets" => "RAILS_MASTER_KEY=456")
+    Kamal::Configuration.any_instance.stubs(:run_id).returns("12345678901234567890123456789012")
 
-    @config = { service: "app", image: "dhh/app", registry: { "username" => "dhh", "password" => "secret" }, servers: { "web" => [ "1.1.1.1" ], "workers" => [ "1.1.1.2" ] }, env: { "secret" => [ "RAILS_MASTER_KEY" ] }, builder: { "arch" => "amd64" } }
+    @config = { service: "app", image: "dhh/app", registry: { "username" => "dhh", "password" => "secret" }, servers: [ "1.1.1.1" ], env: { "secret" => [ "RAILS_MASTER_KEY" ] }, builder: { "arch" => "amd64" } }
   end
 
   teardown do
@@ -13,13 +14,13 @@ class CommandsAppTest < ActiveSupport::TestCase
 
   test "run" do
     assert_equal \
-      "docker run --detach --restart unless-stopped --name app-web-999 --network kamal -e KAMAL_CONTAINER_NAME=\"app-web-999\" -e KAMAL_VERSION=\"999\" --env-file .kamal/apps/app/env/roles/web.env --log-opt max-size=\"10m\" --label service=\"app\" --label role=\"web\" --label destination 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 destination --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 --network kamal --hostname myhost -e KAMAL_CONTAINER_NAME=\"app-web-999\" -e KAMAL_VERSION=\"999\" --env-file .kamal/apps/app/env/roles/web.env --log-opt max-size=\"10m\" --label service=\"app\" --label role=\"web\" --label destination 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 destination --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
 
@@ -27,14 +28,38 @@ class CommandsAppTest < ActiveSupport::TestCase
     @config[:volumes] = [ "/local/path:/container/path" ]
 
     assert_equal \
-      "docker run --detach --restart unless-stopped --name app-web-999 --network kamal -e KAMAL_CONTAINER_NAME=\"app-web-999\" -e KAMAL_VERSION=\"999\" --env-file .kamal/apps/app/env/roles/web.env --log-opt max-size=\"10m\" --volume /local/path:/container/path --label service=\"app\" --label role=\"web\" --label destination 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 destination --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 healthcheck path" do
+    @config[:healthcheck] = { "path" => "/healthz" }
+
+    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/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 destination --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 healthcheck command" do
+    @config[:healthcheck] = { "cmd" => "/bin/up" }
+
+    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 \"(/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 destination --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-specific healthcheck options" do
+    @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\" -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 destination --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 --network kamal -e KAMAL_CONTAINER_NAME=\"app-jobs-999\" -e KAMAL_VERSION=\"999\" --env-file .kamal/apps/app/env/roles/jobs.env --log-opt max-size=\"10m\" --label service=\"app\" --label role=\"jobs\" --label destination --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\" --label destination --mount \"somewhere\" --cap-add dhh/app:999 bin/jobs",
       new_command(role: "jobs", host: "1.1.1.2").run.join(" ")
   end
 
@@ -42,7 +67,7 @@ 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 --network kamal -e KAMAL_CONTAINER_NAME=\"app-web-999\" -e KAMAL_VERSION=\"999\" --env-file .kamal/apps/app/env/roles/web.env --log-driver \"local\" --log-opt max-size=\"100m\" --log-opt max-file=\"3\" --label service=\"app\" --label role=\"web\" --label destination 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 destination --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
 
@@ -51,7 +76,7 @@ class CommandsAppTest < ActiveSupport::TestCase
     @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 --network kamal -e KAMAL_CONTAINER_NAME=\"app-web-999\" -e KAMAL_VERSION=\"999\" --env-file .kamal/apps/app/env/roles/web.env --log-driver \"local\" --log-opt max-size=\"100m\" --log-opt max-file=\"3\" --label service=\"app\" --label role=\"web\" --label destination 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 destination --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
 
@@ -60,7 +85,7 @@ class CommandsAppTest < ActiveSupport::TestCase
     @config[:env]["tags"] = { "tag1" => { "ENV1" => "value1" } }
 
     assert_equal \
-      "docker run --detach --restart unless-stopped --name app-web-999 --network kamal -e KAMAL_CONTAINER_NAME=\"app-web-999\" -e KAMAL_VERSION=\"999\" --env ENV1=\"value1\" --env-file .kamal/apps/app/env/roles/web.env --log-opt max-size=\"10m\" --label service=\"app\" --label role=\"web\" --label destination 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 ENV1=\"value1\" --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 destination --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
 
@@ -83,15 +108,11 @@ class CommandsAppTest < ActiveSupport::TestCase
       new_command.stop.join(" ")
   end
 
-  test "stop with custom drain timeout" do
-    @config[:drain_timeout] = 20
+  test "stop with custom stop wait time" do
+    @config[:stop_wait_time] = 30
     assert_equal \
-      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker stop",
+      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker stop -t 30",
       new_command.stop.join(" ")
-
-    assert_equal \
-      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=workers --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=workers --filter status=running --filter status=restarting' | head -1 | xargs docker stop -t 20",
-      new_command(role: "workers").stop.join(" ")
   end
 
   test "stop with version" do
@@ -113,65 +134,52 @@ class CommandsAppTest < ActiveSupport::TestCase
       new_command.info.join(" ")
   end
 
-  test "deploy" do
-    assert_equal \
-      "docker exec kamal-proxy kamal-proxy deploy app-web --target \"172.1.0.2:80\" --deploy-timeout \"30s\" --drain-timeout \"30s\" --buffer-requests --buffer-responses --log-request-header \"Cache-Control\" --log-request-header \"Last-Modified\" --log-request-header \"User-Agent\"",
-      new_command.deploy(target: "172.1.0.2").join(" ")
-  end
-
-  test "remove" do
-    assert_equal \
-      "docker exec kamal-proxy kamal-proxy remove app-web --target \"172.1.0.2:80\"",
-      new_command.remove(target: "172.1.0.2").join(" ")
-  end
-
-
 
   test "logs" do
     assert_equal \
-      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --timestamps 2>&1",
+      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs 2>&1",
       new_command.logs.join(" ")
   end
 
   test "logs with since" do
     assert_equal \
-      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --timestamps --since 5m 2>&1",
+      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --since 5m 2>&1",
       new_command.logs(since: "5m").join(" ")
   end
 
   test "logs with lines" do
     assert_equal \
-      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --timestamps --tail 100 2>&1",
+      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --tail 100 2>&1",
       new_command.logs(lines: "100").join(" ")
   end
 
   test "logs with since and lines" do
     assert_equal \
-      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --timestamps --since 5m --tail 100 2>&1",
+      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --since 5m --tail 100 2>&1",
       new_command.logs(since: "5m", lines: "100").join(" ")
   end
 
   test "logs with grep" do
     assert_equal \
-      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --timestamps 2>&1 | grep 'my-id'",
+      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs 2>&1 | grep 'my-id'",
       new_command.logs(grep: "my-id").join(" ")
   end
 
   test "logs with grep and grep options" do
     assert_equal \
-      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --timestamps 2>&1 | grep 'my-id' -C 2",
+      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs 2>&1 | grep 'my-id' -C 2",
       new_command.logs(grep: "my-id", grep_options: "-C 2").join(" ")
   end
 
   test "logs with since, grep and grep options" do
     assert_equal \
-      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --timestamps --since 5m 2>&1 | grep 'my-id' -C 2",
+      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --since 5m 2>&1 | grep 'my-id' -C 2",
       new_command.logs(since: "5m", grep: "my-id", grep_options: "-C 2").join(" ")
   end
 
   test "logs with since and grep" do
     assert_equal \
-      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --timestamps --since 5m 2>&1 | grep 'my-id'",
+      "sh -c 'docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''{{.ID}}'\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting' | head -1 | xargs docker logs --since 5m 2>&1 | grep 'my-id'",
       new_command.logs(since: "5m", grep: "my-id").join(" ")
   end
 
@@ -191,22 +199,18 @@ class CommandsAppTest < ActiveSupport::TestCase
     assert_equal \
       "ssh -t root@app-1 -p 22 'sh -c '\\''docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''\\'\\'''\\''{{.ID}}'\\''\\'\\'''\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting'\\'' | head -1 | xargs docker logs --timestamps --tail 123 --follow 2>&1 | grep \"Completed\"'",
       new_command.follow_logs(host: "app-1", lines: 123, grep: "Completed")
-
-    assert_equal \
-      "ssh -t root@app-1 -p 22 'sh -c '\\''docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting --filter ancestor=$(docker image ls --filter reference=dhh/app:latest --format '\\''\\'\\'''\\''{{.ID}}'\\''\\'\\'''\\'') ; docker ps --latest --quiet --filter label=service=app --filter label=role=web --filter status=running --filter status=restarting'\\'' | head -1 | xargs docker logs --tail 123 --follow 2>&1 | grep \"Completed\"'",
-      new_command.follow_logs(host: "app-1", timestamps: false, lines: 123, grep: "Completed")
   end
 
 
   test "execute in new container" do
     assert_equal \
-      "docker run --rm --network kamal --env-file .kamal/apps/app/env/roles/web.env dhh/app:999 bin/rails db:setup",
+      "docker run --rm --env-file .kamal/env/roles/app-web.env dhh/app:999 bin/rails db:setup",
       new_command.execute_in_new_container("bin/rails", "db:setup", env: {}).join(" ")
   end
 
   test "execute in new container with env" do
     assert_equal \
-      "docker run --rm --network kamal --env-file .kamal/apps/app/env/roles/web.env --env foo=\"bar\" dhh/app:999 bin/rails db:setup",
+      "docker run --rm --env-file .kamal/env/roles/app-web.env --env foo=\"bar\" dhh/app:999 bin/rails db:setup",
       new_command.execute_in_new_container("bin/rails", "db:setup", env: { "foo" => "bar" }).join(" ")
   end
 
@@ -215,14 +219,14 @@ class CommandsAppTest < ActiveSupport::TestCase
     @config[:env]["tags"] = { "tag1" => { "ENV1" => "value1" } }
 
     assert_equal \
-      "docker run --rm --network kamal --env ENV1=\"value1\" --env-file .kamal/apps/app/env/roles/web.env dhh/app:999 bin/rails db:setup",
+      "docker run --rm --env ENV1=\"value1\" --env-file .kamal/env/roles/app-web.env dhh/app:999 bin/rails db:setup",
       new_command.execute_in_new_container("bin/rails", "db:setup", env: {}).join(" ")
   end
 
   test "execute in new container with custom options" do
     @config[:servers] = { "web" => { "hosts" => [ "1.1.1.1" ], "options" => { "mount" => "somewhere", "cap-add" => true } } }
     assert_equal \
-      "docker run --rm --network kamal --env-file .kamal/apps/app/env/roles/web.env --mount \"somewhere\" --cap-add dhh/app:999 bin/rails db:setup",
+      "docker run --rm --env-file .kamal/env/roles/app-web.env --mount \"somewhere\" --cap-add dhh/app:999 bin/rails db:setup",
       new_command.execute_in_new_container("bin/rails", "db:setup", env: {}).join(" ")
   end
 
@@ -239,7 +243,7 @@ class CommandsAppTest < ActiveSupport::TestCase
   end
 
   test "execute in new container over ssh" do
-    assert_match %r{docker run -it --rm --network kamal --env-file .kamal/apps/app/env/roles/web.env dhh/app:999 bin/rails c},
+    assert_match %r{docker run -it --rm --env-file .kamal/env/roles/app-web.env dhh/app:999 bin/rails c},
       new_command.execute_in_new_container_over_ssh("bin/rails", "c", env: {})
   end
 
@@ -247,13 +251,13 @@ class CommandsAppTest < ActiveSupport::TestCase
     @config[:servers] = [ { "1.1.1.1" => "tag1" } ]
     @config[:env]["tags"] = { "tag1" => { "ENV1" => "value1" } }
 
-    assert_equal "ssh -t root@1.1.1.1 -p 22 'docker run -it --rm --network kamal --env ENV1=\"value1\" --env-file .kamal/apps/app/env/roles/web.env dhh/app:999 bin/rails c'",
+    assert_equal "ssh -t root@1.1.1.1 -p 22 'docker run -it --rm --env ENV1=\"value1\" --env-file .kamal/env/roles/app-web.env dhh/app:999 bin/rails c'",
       new_command.execute_in_new_container_over_ssh("bin/rails", "c", env: {})
   end
 
   test "execute in new container with custom options over ssh" do
     @config[:servers] = { "web" => { "hosts" => [ "1.1.1.1" ], "options" => { "mount" => "somewhere", "cap-add" => true } } }
-    assert_match %r{docker run -it --rm --network kamal --env-file .kamal/apps/app/env/roles/web.env --mount \"somewhere\" --cap-add dhh/app:999 bin/rails c},
+    assert_match %r{docker run -it --rm --env-file .kamal/env/roles/app-web.env --mount \"somewhere\" --cap-add dhh/app:999 bin/rails c},
       new_command.execute_in_new_container_over_ssh("bin/rails", "c", env: {})
   end
 
@@ -408,34 +412,48 @@ class CommandsAppTest < ActiveSupport::TestCase
       new_command.tag_latest_image.join(" ")
   end
 
+  test "cord" do
+    assert_equal "docker inspect -f '{{ range .Mounts }}{{printf \"%s %s\\n\" .Source .Destination}}{{ end }}' app-web-123 | awk '$2 == \"/tmp/kamal-cord\" {print $1}'", new_command.cord(version: 123).join(" ")
+  end
+
+  test "tie cord" do
+    assert_equal "mkdir -p . ; touch cordfile", new_command.tie_cord("cordfile").join(" ")
+    assert_equal "mkdir -p corddir ; touch corddir/cordfile", new_command.tie_cord("corddir/cordfile").join(" ")
+    assert_equal "mkdir -p /corddir ; touch /corddir/cordfile", new_command.tie_cord("/corddir/cordfile").join(" ")
+  end
+
+  test "cut cord" do
+    assert_equal "rm -r corddir", new_command.cut_cord("corddir").join(" ")
+  end
+
   test "extract assets" do
     assert_equal [
-      :mkdir, "-p", ".kamal/apps/app/assets/extracted/web-999", "&&",
+      :mkdir, "-p", ".kamal/assets/extracted/app-web-999", "&&",
       :docker, :stop, "-t 1", "app-web-assets", "2> /dev/null", "|| true", "&&",
-      :docker, :run, "--name", "app-web-assets", "--detach", "--rm", "--entrypoint", "sleep", "dhh/app:999", "1000000", "&&",
-      :docker, :cp, "-L", "app-web-assets:/public/assets/.", ".kamal/apps/app/assets/extracted/web-999", "&&",
+      :docker, :run, "--name", "app-web-assets", "--detach", "--rm", "dhh/app:999", "sleep 1000000", "&&",
+      :docker, :cp, "-L", "app-web-assets:/public/assets/.", ".kamal/assets/extracted/app-web-999", "&&",
       :docker, :stop, "-t 1", "app-web-assets"
     ], new_command(asset_path: "/public/assets").extract_assets
   end
 
   test "sync asset volumes" do
     assert_equal [
-      :mkdir, "-p", ".kamal/apps/app/assets/volumes/web-999", ";",
-      :cp, "-rnT", ".kamal/apps/app/assets/extracted/web-999", ".kamal/apps/app/assets/volumes/web-999"
+      :mkdir, "-p", ".kamal/assets/volumes/app-web-999", ";",
+      :cp, "-rnT", ".kamal/assets/extracted/app-web-999", ".kamal/assets/volumes/app-web-999"
     ], new_command(asset_path: "/public/assets").sync_asset_volumes
 
     assert_equal [
-      :mkdir, "-p", ".kamal/apps/app/assets/volumes/web-999", ";",
-      :cp, "-rnT", ".kamal/apps/app/assets/extracted/web-999", ".kamal/apps/app/assets/volumes/web-999", ";",
-      :cp, "-rnT", ".kamal/apps/app/assets/extracted/web-999", ".kamal/apps/app/assets/volumes/web-998", "|| true", ";",
-      :cp, "-rnT", ".kamal/apps/app/assets/extracted/web-998", ".kamal/apps/app/assets/volumes/web-999", "|| true"
+      :mkdir, "-p", ".kamal/assets/volumes/app-web-999", ";",
+      :cp, "-rnT", ".kamal/assets/extracted/app-web-999", ".kamal/assets/volumes/app-web-999", ";",
+      :cp, "-rnT", ".kamal/assets/extracted/app-web-999", ".kamal/assets/volumes/app-web-998", "|| true", ";",
+      :cp, "-rnT", ".kamal/assets/extracted/app-web-998", ".kamal/assets/volumes/app-web-999", "|| true"
     ], new_command(asset_path: "/public/assets").sync_asset_volumes(old_version: 998)
   end
 
   test "clean up assets" do
     assert_equal [
-      :find, ".kamal/apps/app/assets/extracted", "-maxdepth 1", "-name", "'web-*'", "!", "-name", "web-999", "-exec rm -rf \"{}\" +", ";",
-      :find, ".kamal/apps/app/assets/volumes", "-maxdepth 1", "-name", "'web-*'", "!", "-name", "web-999", "-exec rm -rf \"{}\" +"
+      :find, ".kamal/assets/extracted", "-maxdepth 1", "-name", "'app-web-*'", "!", "-name", "app-web-999", "-exec rm -rf \"{}\" +", ";",
+      :find, ".kamal/assets/volumes", "-maxdepth 1", "-name", "'app-web-*'", "!", "-name", "app-web-999", "-exec rm -rf \"{}\" +"
     ], new_command(asset_path: "/public/assets").clean_up_assets
   end
 

test/commands/hook_test.rb

@@ -8,7 +8,7 @@ class CommandsHookTest < ActiveSupport::TestCase
 
     @config = {
       service: "app", image: "dhh/app", registry: { "username" => "dhh", "password" => "secret" }, servers: [ "1.1.1.1" ],
-      builder: { "arch" => "amd64" }
+      builder: { "arch" => "amd64" }, traefik: { "args" => { "accesslog.format" => "json", "metrics.prometheus.buckets" => "0.1,0.3,1.2,5.0" } }
     }
 
     @performer = Kamal::Git.email.presence || `whoami`.chomp
@@ -16,34 +16,41 @@ class CommandsHookTest < ActiveSupport::TestCase
   end
 
   test "run" do
-    assert_equal [ ".kamal/hooks/foo" ], new_command.run("foo")
-  end
-
-  test "env" do
-    assert_equal ({
-      "KAMAL_RECORDED_AT" => @recorded_at,
-      "KAMAL_PERFORMER" => @performer,
-      "KAMAL_VERSION" => "123",
-      "KAMAL_SERVICE_VERSION" => "app@123",
-      "KAMAL_SERVICE" => "app"
-    }), new_command.env
+    assert_equal [
+      ".kamal/hooks/foo",
+      { env: {
+        "KAMAL_RECORDED_AT" => @recorded_at,
+        "KAMAL_PERFORMER" => @performer,
+        "KAMAL_VERSION" => "123",
+        "KAMAL_SERVICE_VERSION" => "app@123",
+        "KAMAL_SERVICE" => "app" } }
+    ], new_command.run("foo")
   end
 
   test "run with custom hooks_path" do
-    assert_equal [ "custom/hooks/path/foo" ], new_command(hooks_path: "custom/hooks/path").run("foo")
+    assert_equal [
+      "custom/hooks/path/foo",
+      { env: {
+        "KAMAL_RECORDED_AT" => @recorded_at,
+        "KAMAL_PERFORMER" => @performer,
+        "KAMAL_VERSION" => "123",
+        "KAMAL_SERVICE_VERSION" => "app@123",
+        "KAMAL_SERVICE" => "app" } }
+    ], new_command(hooks_path: "custom/hooks/path").run("foo")
   end
 
-  test "env with secrets" do
+  test "hook with secrets" do
     with_test_secrets("secrets" => "DB_PASSWORD=secret") do
-      assert_equal (
-        {
+      assert_equal [
+        ".kamal/hooks/foo",
+        { env: {
           "KAMAL_RECORDED_AT" => @recorded_at,
           "KAMAL_PERFORMER" => @performer,
           "KAMAL_VERSION" => "123",
           "KAMAL_SERVICE_VERSION" => "app@123",
           "KAMAL_SERVICE" => "app",
-          "DB_PASSWORD" => "secret" }
-      ), new_command.env(secrets: true)
+          "DB_PASSWORD" => "secret" } }
+      ], new_command(env: { "secret" => [ "DB_PASSWORD" ] }).run("foo", secrets: true)
     end
   end
 

test/commands/lock_test.rb

@@ -4,25 +4,25 @@ class CommandsLockTest < ActiveSupport::TestCase
   setup do
     @config = {
       service: "app", image: "dhh/app", registry: { "username" => "dhh", "password" => "secret" }, servers: [ "1.1.1.1" ],
-      builder: { "arch" => "amd64" }
+      builder: { "arch" => "amd64" }, traefik: { "args" => { "accesslog.format" => "json", "metrics.prometheus.buckets" => "0.1,0.3,1.2,5.0" } }
     }
   end
 
   test "status" do
     assert_equal \
-      "stat .kamal/lock-app-production > /dev/null && cat .kamal/lock-app-production/details | base64 -d",
+      "stat .kamal/locks/app-production > /dev/null && cat .kamal/locks/app-production/details | base64 -d",
       new_command.status.join(" ")
   end
 
   test "acquire" do
     assert_match \
-      %r{mkdir \.kamal/lock-app-production && echo ".*" > \.kamal/lock-app-production/details}m,
+      %r{mkdir \.kamal/locks/app-production && echo ".*" > \.kamal/locks/app-production/details}m,
       new_command.acquire("Hello", "123").join(" ")
   end
 
   test "release" do
     assert_match \
-      "rm .kamal/lock-app-production/details && rm -r .kamal/lock-app-production",
+      "rm .kamal/locks/app-production/details && rm -r .kamal/locks/app-production",
       new_command.release.join(" ")
   end
 

test/commands/proxy_test.rb

@@ -15,13 +15,13 @@ class CommandsProxyTest < ActiveSupport::TestCase
 
   test "run" do
     assert_equal \
-      "docker run --name kamal-proxy --network kamal --detach --restart unless-stopped --publish 80:80 --publish 443:443 --volume kamal-proxy-config:/home/kamal-proxy/.config/kamal-proxy --log-opt max-size=\"10m\" basecamp/kamal-proxy:#{Kamal::Configuration::PROXY_MINIMUM_VERSION}",
+      "docker run --name kamal-proxy --network kamal --detach --restart unless-stopped --publish 80:80 --publish 443:443 --volume /var/run/docker.sock:/var/run/docker.sock --volume $(pwd)/.kamal/proxy/config:/root/.config/kamal-proxy --log-opt max-size=\"10m\" #{Kamal::Configuration::Proxy::DEFAULT_IMAGE}",
       new_command.run.join(" ")
   end
 
   test "run with ports configured" do
     assert_equal \
-      "docker run --name kamal-proxy --network kamal --detach --restart unless-stopped --publish 80:80 --publish 443:443 --volume kamal-proxy-config:/home/kamal-proxy/.config/kamal-proxy --log-opt max-size=\"10m\" basecamp/kamal-proxy:#{Kamal::Configuration::PROXY_MINIMUM_VERSION}",
+      "docker run --name kamal-proxy --network kamal --detach --restart unless-stopped --publish 80:80 --publish 443:443 --volume /var/run/docker.sock:/var/run/docker.sock --volume $(pwd)/.kamal/proxy/config:/root/.config/kamal-proxy --log-opt max-size=\"10m\" #{Kamal::Configuration::Proxy::DEFAULT_IMAGE}",
       new_command.run.join(" ")
   end
 
@@ -29,7 +29,7 @@ class CommandsProxyTest < ActiveSupport::TestCase
     @config.delete(:proxy)
 
     assert_equal \
-      "docker run --name kamal-proxy --network kamal --detach --restart unless-stopped --publish 80:80 --publish 443:443 --volume kamal-proxy-config:/home/kamal-proxy/.config/kamal-proxy --log-opt max-size=\"10m\" basecamp/kamal-proxy:#{Kamal::Configuration::PROXY_MINIMUM_VERSION}",
+      "docker run --name kamal-proxy --network kamal --detach --restart unless-stopped --publish 80:80 --publish 443:443 --volume /var/run/docker.sock:/var/run/docker.sock --volume $(pwd)/.kamal/proxy/config:/root/.config/kamal-proxy --log-opt max-size=\"10m\" #{Kamal::Configuration::Proxy::DEFAULT_IMAGE}",
       new_command.run.join(" ")
   end
 
@@ -37,7 +37,7 @@ class CommandsProxyTest < ActiveSupport::TestCase
     @config[:logging] = { "driver" => "local", "options" => { "max-size" => "100m", "max-file" => "3" } }
 
     assert_equal \
-      "docker run --name kamal-proxy --network kamal --detach --restart unless-stopped --publish 80:80 --publish 443:443 --volume kamal-proxy-config:/home/kamal-proxy/.config/kamal-proxy --log-driver \"local\" --log-opt max-size=\"100m\" --log-opt max-file=\"3\" basecamp/kamal-proxy:#{Kamal::Configuration::PROXY_MINIMUM_VERSION}",
+      "docker run --name kamal-proxy --network kamal --detach --restart unless-stopped --publish 80:80 --publish 443:443 --volume /var/run/docker.sock:/var/run/docker.sock --volume $(pwd)/.kamal/proxy/config:/root/.config/kamal-proxy --log-driver \"local\" --log-opt max-size=\"100m\" --log-opt max-file=\"3\" #{Kamal::Configuration::Proxy::DEFAULT_IMAGE}",
       new_command.run.join(" ")
   end
 
@@ -67,22 +67,16 @@ class CommandsProxyTest < ActiveSupport::TestCase
 
   test "proxy logs since 2h" do
     assert_equal \
-      "docker logs kamal-proxy --since 2h --timestamps 2>&1",
+      "docker logs kamal-proxy  --since 2h --timestamps 2>&1",
       new_command.logs(since: "2h").join(" ")
   end
 
   test "proxy logs last 10 lines" do
     assert_equal \
-      "docker logs kamal-proxy --tail 10 --timestamps 2>&1",
+      "docker logs kamal-proxy  --tail 10 --timestamps 2>&1",
       new_command.logs(lines: 10).join(" ")
   end
 
-  test "proxy logs without timestamps" do
-    assert_equal \
-      "docker logs kamal-proxy 2>&1",
-      new_command.logs(timestamps: false).join(" ")
-  end
-
   test "proxy logs with grep hello!" do
     assert_equal \
       "docker logs kamal-proxy --timestamps 2>&1 | grep 'hello!'",
@@ -113,10 +107,16 @@ class CommandsProxyTest < ActiveSupport::TestCase
       new_command.follow_logs(host: @config[:servers].first, grep: "hello!")
   end
 
-  test "version" do
+  test "deploy" do
     assert_equal \
-      "docker inspect kamal-proxy --format '{{.Config.Image}}' | cut -d: -f2",
-      new_command.version.join(" ")
+      "docker exec kamal-proxy kamal-proxy deploy service --target \"172.1.0.2:80\" --buffer-requests --buffer-responses --log-request-header \"Cache-Control\" --log-request-header \"Last-Modified\"",
+      new_command.deploy("service", target: "172.1.0.2").join(" ")
+  end
+
+  test "remove" do
+    assert_equal \
+      "docker exec kamal-proxy kamal-proxy remove service --target \"172.1.0.2:80\"",
+      new_command.remove("service", target: "172.1.0.2").join(" ")
   end
 
   private

test/commands/prune_test.rb

@@ -4,7 +4,7 @@ class CommandsPruneTest < ActiveSupport::TestCase
   setup do
     @config = {
       service: "app", image: "dhh/app", registry: { "username" => "dhh", "password" => "secret" }, servers: [ "1.1.1.1" ],
-      builder: { "arch" => "amd64" }
+      builder: { "arch" => "amd64" }, traefik: { "args" => { "accesslog.format" => "json", "metrics.prometheus.buckets" => "0.1,0.3,1.2,5.0" } }
     }
   end
 
@@ -30,6 +30,12 @@ class CommandsPruneTest < ActiveSupport::TestCase
       new_command.app_containers(retain: 3).join(" ")
   end
 
+  test "healthcheck containers" do
+    assert_equal \
+      "docker container prune --force --filter label=service=healthcheck-app",
+      new_command.healthcheck_containers.join(" ")
+  end
+
   private
     def new_command
       Kamal::Commands::Prune.new(Kamal::Configuration.new(@config, version: "123"))

test/commands/registry_test.rb

@@ -55,15 +55,6 @@ class CommandsRegistryTest < ActiveSupport::TestCase
       registry.logout.join(" ")
   end
 
-  test "registry setup" do
-    @config[:registry] = { "server" => "localhost:5000" }
-    assert_equal "docker start kamal-docker-registry || docker run --detach -p 5000:5000 --name kamal-docker-registry registry:2", registry.setup.join(" ")
-  end
-
-  test "registry remove" do
-    assert_equal "docker stop kamal-docker-registry && docker rm kamal-docker-registry", registry.remove.join(" ")
-  end
-
   private
     def registry
       Kamal::Commands::Registry.new Kamal::Configuration.new(@config)

test/commands/server_test.rb

@@ -4,7 +4,7 @@ class CommandsServerTest < ActiveSupport::TestCase
   setup do
     @config = {
       service: "app", image: "dhh/app", registry: { "username" => "dhh", "password" => "secret" }, servers: [ "1.1.1.1" ],
-      builder: { "arch" => "amd64" }
+      builder: { "arch" => "amd64" }, traefik: { "args" => { "accesslog.format" => "json", "metrics.prometheus.buckets" => "0.1,0.3,1.2,5.0" } }
     }
   end
 

test/commands/traefik_test.rb

@@ -0,0 +1,195 @@
+require "test_helper"
+
+class CommandsTraefikTest < ActiveSupport::TestCase
+  setup do
+    @image = "traefik:test"
+
+    @config = {
+      service: "app", image: "dhh/app", registry: { "username" => "dhh", "password" => "secret" }, servers: [ "1.1.1.1" ], builder: { "arch" => "amd64" },
+      traefik: { "image" => @image, "args" => { "accesslog.format" => "json", "api.insecure" => true, "metrics.prometheus.buckets" => "0.1,0.3,1.2,5.0" } }
+    }
+
+    setup_test_secrets("secrets" => "EXAMPLE_API_KEY=456")
+  end
+
+  teardown do
+    teardown_test_secrets
+  end
+
+  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\" --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\" --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\" --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\" --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\" --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\" --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\" --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\" --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\" --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\" --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.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\" --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"] = { "EXAMPLE_API_KEY" => "456" }
+    assert_equal \
+      "docker run --name traefik --detach --restart unless-stopped --publish 80:80 --volume /var/run/docker.sock:/var/run/docker.sock --env EXAMPLE_API_KEY=\"456\" --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 without configuration" do
+    @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\" --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::Configuration::Traefik::DEFAULT_IMAGE} --providers.docker --log.level=\"DEBUG\"",
+      new_command.run.join(" ")
+  end
+
+  test "run with logging config" do
+    @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\" --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 default args overriden" do
+    @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\" --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
+
+  test "run with args array" do
+    @config[:traefik]["args"] = { "entrypoints.web.forwardedheaders.trustedips" => %w[ 127.0.0.1 127.0.0.2 ] }
+    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.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\" traefik:test --providers.docker --log.level=\"DEBUG\" --entrypoints.web.forwardedheaders.trustedips=\"127.0.0.1\" --entrypoints.web.forwardedheaders.trustedips=\"127.0.0.2\"", new_command.run.join(" ")
+  end
+
+  test "traefik start" do
+    assert_equal \
+      "docker container start traefik",
+      new_command.start.join(" ")
+  end
+
+  test "traefik stop" do
+    assert_equal \
+      "docker container stop traefik",
+      new_command.stop.join(" ")
+  end
+
+  test "traefik info" do
+    assert_equal \
+      "docker ps --filter name=^traefik$",
+      new_command.info.join(" ")
+  end
+
+  test "traefik logs" do
+    assert_equal \
+      "docker logs traefik --timestamps 2>&1",
+      new_command.logs.join(" ")
+  end
+
+  test "traefik logs since 2h" do
+    assert_equal \
+      "docker logs traefik  --since 2h --timestamps 2>&1",
+      new_command.logs(since: "2h").join(" ")
+  end
+
+  test "traefik logs last 10 lines" do
+    assert_equal \
+      "docker logs traefik  --tail 10 --timestamps 2>&1",
+      new_command.logs(lines: 10).join(" ")
+  end
+
+  test "traefik logs with grep hello!" do
+    assert_equal \
+      "docker logs traefik --timestamps 2>&1 | grep 'hello!'",
+      new_command.logs(grep: "hello!").join(" ")
+  end
+
+  test "traefik logs with grep hello! and grep options" do
+    assert_equal \
+      "docker logs traefik --timestamps 2>&1 | grep 'hello!' -C 2",
+      new_command.logs(grep: "hello!", grep_options: "-C 2").join(" ")
+  end
+
+  test "traefik remove container" do
+    assert_equal \
+      "docker container prune --force --filter label=org.opencontainers.image.title=Traefik",
+      new_command.remove_container.join(" ")
+  end
+
+  test "traefik remove image" do
+    assert_equal \
+      "docker image prune --all --force --filter label=org.opencontainers.image.title=Traefik",
+      new_command.remove_image.join(" ")
+  end
+
+  test "traefik follow logs" do
+    assert_equal \
+      "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 -p 22 'docker logs traefik --timestamps --tail 10 --follow 2>&1 | grep \"hello!\"'",
+      new_command.follow_logs(host: @config[:servers].first, grep: "hello!")
+  end
+
+  private
+    def new_command
+      Kamal::Commands::Traefik.new(Kamal::Configuration.new(@config, version: "123"))
+    end
+end

test/configuration/accessory_test.rb

@@ -119,9 +119,9 @@ class ConfigurationAccessoryTest < ActiveSupport::TestCase
     with_test_secrets("secrets" => "MYSQL_ROOT_PASSWORD=secret123") do
       config = Kamal::Configuration.new(@deploy)
 
-      assert_equal [ "--env", "MYSQL_ROOT_HOST=\"%\"", "--env-file", ".kamal/apps/app/env/accessories/mysql.env" ], config.accessory(:mysql).env_args.map(&:to_s)
+      assert_equal [ "--env", "MYSQL_ROOT_HOST=\"%\"", "--env-file", ".kamal/env/accessories/app-mysql.env" ], config.accessory(:mysql).env_args.map(&:to_s)
       assert_equal "MYSQL_ROOT_PASSWORD=secret123\n", config.accessory(:mysql).secrets_io.string
-      assert_equal [ "--env", "SOMETHING=\"else\"", "--env-file", ".kamal/apps/app/env/accessories/redis.env" ], @config.accessory(:redis).env_args
+      assert_equal [ "--env", "SOMETHING=\"else\"", "--env-file", ".kamal/env/accessories/app-redis.env" ], @config.accessory(:redis).env_args
       assert_equal "\n", config.accessory(:redis).secrets_io.string
     end
   end

test/configuration/role_test.rb

@@ -39,7 +39,7 @@ class ConfigurationRoleTest < ActiveSupport::TestCase
   end
 
   test "special label args for web" do
-    assert_equal [ "--label", "service=\"app\"", "--label", "role=\"web\"", "--label", "destination" ], config.role(:web).label_args
+    assert_equal [ "--label", "service=\"app\"", "--label", "role=\"web\"", "--label", "destination", "--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
@@ -53,19 +53,24 @@ class ConfigurationRoleTest < ActiveSupport::TestCase
     assert_equal "70", Kamal::Configuration.new(@deploy_with_roles).role(:workers).labels["my.custom.label"]
   end
 
-  test "default proxy label on non-web role" do
+  test "overwriting default traefik label" do
+    @deploy[:labels] = { "traefik.http.routers.app-web.rule" => "\"Host(\\`example.com\\`) || (Host(\\`example.org\\`) && Path(\\`/traefik\\`))\"" }
+    assert_equal "\"Host(\\`example.com\\`) || (Host(\\`example.org\\`) && Path(\\`/traefik\\`))\"", config.role(:web).labels["traefik.http.routers.app-web.rule"]
+  end
+
+  test "default traefik label on non-web role" do
     config = Kamal::Configuration.new(@deploy_with_roles.tap { |c|
-      c[:servers]["beta"] = { "proxy" => true, "hosts" => [ "1.1.1.5" ] }
+      c[:servers]["beta"] = { "traefik" => true, "hosts" => [ "1.1.1.5" ] }
     })
 
-    assert_equal [ "--label", "service=\"app\"", "--label", "role=\"beta\"", "--label", "destination" ], config.role(:beta).label_args
+    assert_equal [ "--label", "service=\"app\"", "--label", "role=\"beta\"", "--label", "destination", "--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
     assert_equal "redis://a/b", config_with_roles.role(:workers).env("1.1.1.3").clear["REDIS_URL"]
 
     assert_equal \
-      [ "--env", "REDIS_URL=\"redis://a/b\"", "--env", "WEB_CONCURRENCY=\"4\"", "--env-file", ".kamal/apps/app/env/roles/workers.env" ],
+      [ "--env", "REDIS_URL=\"redis://a/b\"", "--env", "WEB_CONCURRENCY=\"4\"", "--env-file", ".kamal/env/roles/app-workers.env" ],
       config_with_roles.role(:workers).env_args("1.1.1.3").map(&:to_s)
 
     assert_equal \
@@ -84,7 +89,7 @@ class ConfigurationRoleTest < ActiveSupport::TestCase
 
   test "env args" do
     assert_equal \
-      [ "--env", "REDIS_URL=\"redis://a/b\"", "--env", "WEB_CONCURRENCY=\"4\"", "--env-file", ".kamal/apps/app/env/roles/workers.env" ],
+      [ "--env", "REDIS_URL=\"redis://a/b\"", "--env", "WEB_CONCURRENCY=\"4\"", "--env-file", ".kamal/env/roles/app-workers.env" ],
       config_with_roles.role(:workers).env_args("1.1.1.3").map(&:to_s)
 
     assert_equal \
@@ -114,7 +119,7 @@ class ConfigurationRoleTest < ActiveSupport::TestCase
       }
 
       assert_equal \
-        [ "--env", "REDIS_URL=\"redis://a/b\"", "--env", "WEB_CONCURRENCY=\"4\"", "--env-file", ".kamal/apps/app/env/roles/workers.env" ],
+        [ "--env", "REDIS_URL=\"redis://a/b\"", "--env", "WEB_CONCURRENCY=\"4\"", "--env-file", ".kamal/env/roles/app-workers.env" ],
         config_with_roles.role(:workers).env_args("1.1.1.3").map(&:to_s)
 
       assert_equal \
@@ -136,7 +141,7 @@ class ConfigurationRoleTest < ActiveSupport::TestCase
       }
 
       assert_equal \
-        [ "--env", "REDIS_URL=\"redis://a/b\"", "--env", "WEB_CONCURRENCY=\"4\"", "--env-file", ".kamal/apps/app/env/roles/workers.env" ],
+        [ "--env", "REDIS_URL=\"redis://a/b\"", "--env", "WEB_CONCURRENCY=\"4\"", "--env-file", ".kamal/env/roles/app-workers.env" ],
         config_with_roles.role(:workers).env_args("1.1.1.3").map(&:to_s)
 
       assert_equal \
@@ -157,7 +162,7 @@ class ConfigurationRoleTest < ActiveSupport::TestCase
       }
 
       assert_equal \
-        [ "--env", "REDIS_URL=\"redis://a/b\"", "--env", "WEB_CONCURRENCY=\"4\"", "--env-file", ".kamal/apps/app/env/roles/workers.env" ],
+        [ "--env", "REDIS_URL=\"redis://a/b\"", "--env", "WEB_CONCURRENCY=\"4\"", "--env-file", ".kamal/env/roles/app-workers.env" ],
         config_with_roles.role(:workers).env_args("1.1.1.3").map(&:to_s)
 
       assert_equal \
@@ -184,7 +189,7 @@ class ConfigurationRoleTest < ActiveSupport::TestCase
       }
 
       assert_equal \
-        [ "--env", "REDIS_URL=\"redis://c/d\"", "--env-file", ".kamal/apps/app/env/roles/workers.env" ],
+        [ "--env", "REDIS_URL=\"redis://c/d\"", "--env-file", ".kamal/env/roles/app-workers.env" ],
         config_with_roles.role(:workers).env_args("1.1.1.3").map(&:to_s)
 
       assert_equal \
@@ -193,6 +198,26 @@ class ConfigurationRoleTest < ActiveSupport::TestCase
     end
   end
 
+  test "uses cord" do
+    assert config_with_roles.role(:web).uses_cord?
+    assert_not config_with_roles.role(:workers).uses_cord?
+  end
+
+  test "cord host file" do
+    assert_match %r{.kamal/cords/app-web-[0-9a-f]{32}/cord}, config_with_roles.role(:web).cord_host_file
+  end
+
+  test "cord volume" do
+    assert_equal "/tmp/kamal-cord", config_with_roles.role(:web).cord_volume.container_path
+    assert_match %r{.kamal/cords/app-web-[0-9a-f]{32}}, config_with_roles.role(:web).cord_volume.host_path
+    assert_equal "--volume", config_with_roles.role(:web).cord_volume.docker_args[0]
+    assert_match %r{\$\(pwd\)/.kamal/cords/app-web-[0-9a-f]{32}:/tmp/kamal-cord}, config_with_roles.role(:web).cord_volume.docker_args[1]
+  end
+
+  test "cord container file" do
+    assert_equal "/tmp/kamal-cord/cord", config_with_roles.role(:web).cord_container_file
+  end
+
   test "asset path and volume args" do
     ENV["VERSION"] = "12345"
     assert_nil config_with_roles.role(:web).asset_volume_args
@@ -207,7 +232,7 @@ class ConfigurationRoleTest < ActiveSupport::TestCase
     })
     assert_equal "foo", config_with_assets.role(:web).asset_path
     assert_equal "foo", config_with_assets.role(:workers).asset_path
-    assert_equal [ "--volume", "$(pwd)/.kamal/apps/app/assets/volumes/web-12345:foo" ], config_with_assets.role(:web).asset_volume_args
+    assert_equal [ "--volume", "$(pwd)/.kamal/assets/volumes/app-web-12345:foo" ], config_with_assets.role(:web).asset_volume_args
     assert_nil config_with_assets.role(:workers).asset_volume_args
     assert config_with_assets.role(:web).assets?
     assert_not config_with_assets.role(:workers).assets?
@@ -217,7 +242,7 @@ class ConfigurationRoleTest < ActiveSupport::TestCase
     })
     assert_equal "bar", config_with_assets.role(:web).asset_path
     assert_nil config_with_assets.role(:workers).asset_path
-    assert_equal [ "--volume", "$(pwd)/.kamal/apps/app/assets/volumes/web-12345:bar" ], config_with_assets.role(:web).asset_volume_args
+    assert_equal [ "--volume", "$(pwd)/.kamal/assets/volumes/app-web-12345:bar" ], config_with_assets.role(:web).asset_volume_args
     assert_nil config_with_assets.role(:workers).asset_volume_args
     assert config_with_assets.role(:web).assets?
     assert_not config_with_assets.role(:workers).assets?
@@ -228,36 +253,20 @@ class ConfigurationRoleTest < ActiveSupport::TestCase
 
   test "asset extracted path" do
     ENV["VERSION"] = "12345"
-    assert_equal ".kamal/apps/app/assets/extracted/web-12345", config_with_roles.role(:web).asset_extracted_directory
-    assert_equal ".kamal/apps/app/assets/extracted/workers-12345", config_with_roles.role(:workers).asset_extracted_directory
+    assert_equal ".kamal/assets/extracted/app-web-12345", config_with_roles.role(:web).asset_extracted_path
+    assert_equal ".kamal/assets/extracted/app-workers-12345", config_with_roles.role(:workers).asset_extracted_path
   ensure
     ENV.delete("VERSION")
   end
 
   test "asset volume path" do
     ENV["VERSION"] = "12345"
-    assert_equal ".kamal/apps/app/assets/volumes/web-12345", config_with_roles.role(:web).asset_volume_directory
-    assert_equal ".kamal/apps/app/assets/volumes/workers-12345", config_with_roles.role(:workers).asset_volume_directory
+    assert_equal ".kamal/assets/volumes/app-web-12345", config_with_roles.role(:web).asset_volume_path
+    assert_equal ".kamal/assets/volumes/app-workers-12345", config_with_roles.role(:workers).asset_volume_path
   ensure
     ENV.delete("VERSION")
   end
 
-  test "stop args with proxy" do
-    assert_equal [], config_with_roles.role(:web).stop_args
-  end
-
-  test "stop args with no proxy" do
-    assert_equal [ "-t", 30 ], config_with_roles.role(:workers).stop_args
-  end
-
-  test "role specific proxy config" do
-    @deploy_with_roles[:proxy] = { "response_timeout" => 15 }
-    @deploy_with_roles[:servers]["workers"]["proxy"] = { "response_timeout" => 18 }
-
-    assert_equal "15s", config_with_roles.role(:web).proxy.deploy_options[:"target-timeout"]
-    assert_equal "18s", config_with_roles.role(:workers).proxy.deploy_options[:"target-timeout"]
-  end
-
   private
     def config
       Kamal::Configuration.new(@deploy)

test/configuration/validation_test.rb

@@ -14,7 +14,7 @@ class ConfigurationValidationTest < ActiveSupport::TestCase
       assert_error "#{key}: should be a boolean", **{ key => "foo" }
     end
 
-    [ :deploy_timeout, :drain_timeout, :retain_containers, :readiness_delay ].each do |key|
+    [ :stop_wait_time, :retain_containers, :readiness_delay ].each do |key|
       assert_error "#{key}: should be an integer", **{ key => "foo" }
     end
 
@@ -22,7 +22,7 @@ class ConfigurationValidationTest < ActiveSupport::TestCase
 
     assert_error "servers: should be an array or a hash", servers: "foo"
 
-    [ :labels, :registry, :accessories, :env, :ssh, :sshkit, :builder, :proxy, :boot, :logging ].each do |key|
+    [ :labels, :registry, :accessories, :env, :ssh, :sshkit, :builder, :traefik, :boot, :healthcheck, :logging ].each do |key|
       assert_error "#{key}: should be a hash", **{ key =>[] }
     end
   end

test/configuration_test.rb

@@ -77,22 +77,22 @@ class ConfigurationTest < ActiveSupport::TestCase
     assert_equal "1.1.1.1", @config_with_roles.primary_host
   end
 
-  test "proxy hosts" do
-    assert_equal [ "1.1.1.1", "1.1.1.2" ], @config_with_roles.proxy_hosts
+  test "traefik hosts" do
+    assert_equal [ "1.1.1.1", "1.1.1.2" ], @config_with_roles.traefik_hosts
 
-    @deploy_with_roles[:servers]["workers"]["proxy"] = true
+    @deploy_with_roles[:servers]["workers"]["traefik"] = true
     config = Kamal::Configuration.new(@deploy_with_roles)
 
-    assert_equal [ "1.1.1.1", "1.1.1.2", "1.1.1.3" ], config.proxy_hosts
+    assert_equal [ "1.1.1.1", "1.1.1.2", "1.1.1.3" ], config.traefik_hosts
   end
 
-  test "filtered proxy hosts" do
-    assert_equal [ "1.1.1.1", "1.1.1.2" ], @config_with_roles.proxy_hosts
+  test "filtered traefik hosts" do
+    assert_equal [ "1.1.1.1", "1.1.1.2" ], @config_with_roles.traefik_hosts
 
-    @deploy_with_roles[:servers]["workers"]["proxy"] = true
+    @deploy_with_roles[:servers]["workers"]["traefik"] = true
     config = Kamal::Configuration.new(@deploy_with_roles)
 
-    assert_equal [ "1.1.1.1", "1.1.1.2", "1.1.1.3" ], config.proxy_hosts
+    assert_equal [ "1.1.1.1", "1.1.1.2", "1.1.1.3" ], config.traefik_hosts
   end
 
   test "version no git repo" do
@@ -157,6 +157,10 @@ class ConfigurationTest < ActiveSupport::TestCase
     assert_equal "app-missing", @config.service_with_version
   end
 
+  test "healthcheck service" do
+    assert_equal "healthcheck-app", @config.healthcheck_service
+  end
+
   test "hosts required for all roles" do
     # Empty server list for implied web role
     assert_raises(Kamal::ConfigurationError) do
@@ -265,7 +269,8 @@ class ConfigurationTest < ActiveSupport::TestCase
         sshkit: {},
         volume_args: [ "--volume", "/local/path:/container/path" ],
         builder: { "arch" => "amd64" },
-        logging: [ "--log-opt", "max-size=\"10m\"" ] }
+        logging: [ "--log-opt", "max-size=\"10m\"" ],
+        healthcheck: { "cmd"=>"curl -f http://localhost:3000/up || exit 1", "interval" => "1s", "path"=>"/up", "port"=>3000, "max_attempts" => 7, "cord" => "/tmp/kamal-cord", "log_lines" => 50 } }
 
     assert_equal expected_config, @config.to_h
   end
@@ -291,6 +296,16 @@ class ConfigurationTest < ActiveSupport::TestCase
     assert_equal ".kamal", config.run_directory
   end
 
+  test "run directory as docker volume" do
+    config = Kamal::Configuration.new(@deploy)
+    assert_equal "$(pwd)/.kamal", config.run_directory_as_docker_volume
+  end
+
+  test "run id" do
+    SecureRandom.expects(:hex).with(16).returns("09876543211234567890098765432112")
+    assert_equal "09876543211234567890098765432112", @config.run_id
+  end
+
   test "asset path" do
     assert_nil @config.asset_path
     assert_equal "foo", Kamal::Configuration.new(@deploy.merge!(asset_path: "foo")).asset_path
@@ -307,7 +322,7 @@ class ConfigurationTest < ActiveSupport::TestCase
     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_proxy?
+    assert config.role(:alternate_web).running_traefik?
   end
 
   test "primary role missing" do
@@ -329,52 +344,7 @@ class ConfigurationTest < ActiveSupport::TestCase
     dest_config_file = Pathname.new(File.expand_path("fixtures/deploy_with_extensions.yml", __dir__))
 
     config = Kamal::Configuration.create_from config_file: dest_config_file
-    assert_equal config.role(:web_tokyo).running_proxy?, true
-    assert_equal config.role(:web_chicago).running_proxy?, true
-  end
-
-  test "traefik hooks raise error" do
-    Dir.mktmpdir do |dir|
-      Dir.chdir(dir) do
-        FileUtils.mkdir_p ".kamal/hooks"
-        FileUtils.touch ".kamal/hooks/post-traefik-reboot"
-        FileUtils.touch ".kamal/hooks/pre-traefik-reboot"
-        exception = assert_raises(Kamal::ConfigurationError) do
-          Kamal::Configuration.new(@deploy)
-        end
-        assert_equal "Found pre-traefik-reboot, post-traefik-reboot, these should be renamed to (pre|post)-proxy-reboot", exception.message
-      end
-    end
-  end
-
-  test "proxy ssl roles with no host" do
-    @deploy_with_roles[:servers]["workers"]["proxy"] = { "ssl" => true }
-
-    exception = assert_raises(Kamal::ConfigurationError) do
-      Kamal::Configuration.new(@deploy_with_roles)
-    end
-
-    assert_equal "servers/workers/proxy: Must set a host to enable automatic SSL", exception.message
-  end
-
-  test "proxy ssl roles with multiple servers" do
-    @deploy_with_roles[:servers]["workers"]["proxy"] = { "ssl" => true, "host" => "foo.example.com" }
-
-    exception = assert_raises(Kamal::ConfigurationError) do
-      Kamal::Configuration.new(@deploy_with_roles)
-    end
-
-    assert_equal "SSL is only supported on a single server, found 2 servers for role workers", exception.message
-  end
-
-  test "two proxy ssl roles with same host" do
-    @deploy_with_roles[:servers]["web"] = { "hosts" => [ "1.1.1.1" ], "proxy" => { "ssl" => true, "host" => "foo.example.com" } }
-    @deploy_with_roles[:servers]["workers"] = { "hosts" => [ "1.1.1.1" ], "proxy" => { "ssl" => true, "host" => "foo.example.com" } }
-
-    exception = assert_raises(Kamal::ConfigurationError) do
-      Kamal::Configuration.new(@deploy_with_roles)
-    end
-
-    assert_equal "Different roles can't share the same host for SSL: foo.example.com", exception.message
+    assert_equal config.role(:web_tokyo).running_traefik?, true
+    assert_equal config.role(:web_chicago).running_traefik?, true
   end
 end

test/fixtures/deploy_primary_web_role_override.yml

@@ -2,12 +2,12 @@ service: app
 image: dhh/app
 servers:
   web_chicago:
-    proxy: {}
+    traefik: true
     hosts:
       - 1.1.1.1
       - 1.1.1.2
   web_tokyo:
-    proxy: {}
+    traefik: true
     hosts:
       - 1.1.1.3
       - 1.1.1.4

test/fixtures/deploy_with_extensions.yml

@@ -1,6 +1,6 @@
 
 x-web: &web
-  proxy: {}
+  traefik: true
 
 service: app
 image: dhh/app

test/fixtures/deploy_with_local_registry.yml

@@ -1,37 +0,0 @@
-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:
-  server: localhost:5000
-builder:
-  arch: amd64
-
-accessories:
-  mysql:
-    image: mysql:5.7
-    host: 1.1.1.3
-    port: 3306
-    env:
-      clear:
-        MYSQL_ROOT_HOST: '%'
-      secret:
-        - MYSQL_ROOT_PASSWORD
-    files:
-      - test/fixtures/files/my.cnf:/etc/mysql/my.cnf
-    directories:
-      - data:/var/lib/mysql
-  redis:
-    image: redis:latest
-    roles:
-      - web
-    port: 6379
-    directories:
-      - data:/data
-
-readiness_delay: 0

test/fixtures/deploy_with_multiple_traefik_roles.yml

@@ -8,14 +8,14 @@ servers:
       - 1.1.1.2
     env:
       ROLE: "web"
-    proxy: true
+    traefik: true
   web_tokyo:
     hosts:
       - 1.1.1.3
       - 1.1.1.4
     env:
       ROLE: "web"
-    proxy: true
+    traefik: true
   workers:
     cmd: bin/jobs
     hosts:

test/fixtures/deploy_with_proxy.yml

@@ -13,6 +13,11 @@ registry:
 builder:
   arch: amd64
 
+proxy:
+  enabled: true
+  hosts:
+    - "1.1.1.1"
+  deploy_timeout: 6s
 
 accessories:
   mysql:
@@ -37,4 +42,3 @@ accessories:
       - data:/data
 
 readiness_delay: 0
-deploy_timeout: 6

test/fixtures/deploy_with_proxy_roles.yml

@@ -1,46 +0,0 @@
-service: app
-image: dhh/app
-servers:
-  web:
-    hosts:
-      - "1.1.1.1"
-      - "1.1.1.2"
-  web2:
-    hosts:
-      - "1.1.1.3"
-      - "1.1.1.4"
-    proxy:
-      response_timeout: 15
-registry:
-  username: user
-  password: pw
-builder:
-  arch: amd64
-
-proxy:
-  response_timeout: 10
-
-accessories:
-  mysql:
-    image: mysql:5.7
-    host: 1.1.1.3
-    port: 3306
-    env:
-      clear:
-        MYSQL_ROOT_HOST: '%'
-      secret:
-        - MYSQL_ROOT_PASSWORD
-    files:
-      - test/fixtures/files/my.cnf:/etc/mysql/my.cnf
-    directories:
-      - data:/var/lib/mysql
-  redis:
-    image: redis:latest
-    roles:
-      - web
-    port: 6379
-    directories:
-      - data:/data
-
-readiness_delay: 0
-deploy_timeout: 6

test/fixtures/deploy_with_roles.yml

@@ -16,4 +16,3 @@ registry:
   password: pw
 builder:
   arch: amd64
-deploy_timeout: 1

test/fixtures/deploy_workers_only.yml

@@ -2,7 +2,7 @@ service: app
 image: dhh/app
 servers:
   workers:
-    proxy: false
+    traefik: false
     hosts:
       - 1.1.1.1
       - 1.1.1.2

test/integration/app_test.rb

@@ -20,35 +20,34 @@ class AppTest < IntegrationTest
     wait_for_app_to_be_up
 
     logs = kamal :app, :logs, capture: true
-    assert_match "App Host: vm1", logs
-    assert_match "App Host: vm2", logs
-    assert_match "GET /version HTTP/1.1", logs
+    assert_match /App Host: vm1/, logs
+    assert_match /App Host: vm2/, logs
+    assert_match /GET \/ HTTP\/1.1/, logs
 
     images = kamal :app, :images, capture: true
-    assert_match "App Host: vm1", images
-    assert_match "App Host: vm2", images
-    assert_match /localhost:5000\/app\s+#{latest_app_version}/, images
-    assert_match /localhost:5000\/app\s+latest/, images
+    assert_match /App Host: vm1/, images
+    assert_match /App Host: vm2/, images
+    assert_match /registry:4443\/app\s+#{latest_app_version}/, images
+    assert_match /registry:4443\/app\s+latest/, images
 
     containers = kamal :app, :containers, capture: true
-    assert_match "App Host: vm1", containers
-    assert_match "App Host: vm2", containers
-    assert_match "localhost:5000/app:#{latest_app_version}", containers
-    assert_match "localhost:5000/app:latest", containers
+    assert_match /App Host: vm1/, containers
+    assert_match /App Host: vm2/, containers
+    assert_match /registry:4443\/app:#{latest_app_version}/, containers
+    assert_match /registry:4443\/app:latest/, containers
 
     exec_output = kamal :app, :exec, :ps, capture: true
-    assert_match "App Host: vm1", exec_output
-    assert_match "App Host: vm2", exec_output
+    assert_match /App Host: vm1/, exec_output
+    assert_match /App Host: vm2/, exec_output
     assert_match /1 root      0:\d\d ps/, exec_output
 
     exec_output = kamal :app, :exec, "--reuse", :ps, capture: true
-    assert_match "App Host: vm2", exec_output
-    assert_match "App Host: vm1", exec_output
+    assert_match /App Host: vm1/, exec_output
+    assert_match /App Host: vm2/, exec_output
     assert_match /1 root      0:\d\d nginx/, exec_output
 
     kamal :app, :remove
 
     assert_app_is_down
-    assert_app_directory_removed
   end
 end

test/integration/broken_deploy_test.rb

@@ -9,7 +9,7 @@ class BrokenDeployTest < IntegrationTest
     kamal :deploy
 
     assert_app_is_up version: first_version
-    assert_container_running host: :vm3, name: "app_with_roles-workers-#{first_version}"
+    assert_container_running host: :vm3, name: "app-workers-#{first_version}"
 
     second_version = break_app
 
@@ -17,8 +17,8 @@ class BrokenDeployTest < IntegrationTest
 
     assert_failed_deploy output
     assert_app_is_up version: first_version
-    assert_container_running host: :vm3, name: "app_with_roles-workers-#{first_version}"
-    assert_container_not_running host: :vm3, name: "app_with_roles-workers-#{second_version}"
+    assert_container_running host: :vm3, name: "app-workers-#{first_version}"
+    assert_container_not_running host: :vm3, name: "app-workers-#{second_version}"
   end
 
   private

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

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

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-proxy-reboot

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

@@ -20,18 +20,24 @@ env:
       secret:
         - SECRET_TAG
 asset_path: /usr/share/nginx/html/versions
-deploy_timeout: 2
-drain_timeout: 2
-readiness_delay: 0
-proxy:
-  host: 127.0.0.1
+
 registry:
-  server: localhost:5000
+  server: registry:4443
+  username: root
+  password: root
 builder:
   driver: docker
   arch: <%= Kamal::Utils.docker_arch %>
   args:
     COMMIT_SHA: <%= `git rev-parse HEAD` %>
+healthcheck:
+  cmd: wget -qO- http://localhost > /dev/null || exit 1
+  max_attempts: 3
+traefik:
+  args:
+    accesslog: true
+    accesslog.format: json
+  image: registry:4443/traefik:v2.10
 accessories:
   busybox:
     service: custom-busybox
@@ -39,3 +45,5 @@ accessories:
     cmd: sh -c 'echo "Starting busybox..."; trap exit term; while true; do sleep 1; done'
     roles:
       - web
+stop_wait_time: 1
+readiness_delay: 0

test/integration/docker/deployer/app_with_roles/.kamal/hooks/post-proxy-reboot

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