Start before stopping and longer timeouts

This commit is contained in:
David Heinemeier Hansson
2023-02-22 19:04:23 +01:00
parent b879412a6f
commit 371f98d67f
8 changed files with 45 additions and 18 deletions

View File

@@ -10,13 +10,17 @@ class Mrsk::Cli::App < Mrsk::Cli::Base
execute *MRSK.auditor.record("Booted app version #{version}"), verbosity: :debug execute *MRSK.auditor.record("Booted app version #{version}"), verbosity: :debug
begin begin
execute *MRSK.app.stop, raise_on_non_zero_exit: false old_version = capture_with_info(*MRSK.app.current_running_version).strip
execute *MRSK.app.run(role: role.name) execute *MRSK.app.run(role: role.name)
sleep 10
execute *MRSK.app.stop(version: old_version), raise_on_non_zero_exit: false if old_version.present?
rescue SSHKit::Command::Failed => e rescue SSHKit::Command::Failed => e
if e.message =~ /already in use/ if e.message =~ /already in use/
error "Rebooting container with same version already deployed on #{host}" error "Rebooting container with same version #{version} already deployed on #{host}"
execute *MRSK.auditor.record("Rebooted app version #{version}"), verbosity: :debug execute *MRSK.auditor.record("Rebooted app version #{version}"), verbosity: :debug
execute *MRSK.app.stop(version: version)
execute *MRSK.app.remove_container(version: version) execute *MRSK.app.remove_container(version: version)
execute *MRSK.app.run(role: role.name) execute *MRSK.app.run(role: role.name)
else else

View File

@@ -1,5 +1,5 @@
class Mrsk::Cli::Healthcheck < Mrsk::Cli::Base class Mrsk::Cli::Healthcheck < Mrsk::Cli::Base
MAX_ATTEMPTS = 5 MAX_ATTEMPTS = 7
class HealthcheckError < StandardError; end class HealthcheckError < StandardError; end

View File

@@ -55,11 +55,14 @@ class Mrsk::Cli::Main < Mrsk::Cli::Base
MRSK.version = version MRSK.version = version
if container_name_available?(MRSK.config.service_with_version) if container_name_available?(MRSK.config.service_with_version)
say "Stop current version, then start version #{version}...", :magenta say "Start version #{version}, then stop the old version...", :magenta
on(MRSK.hosts) do |host| on(MRSK.hosts) do |host|
execute *MRSK.app.stop, raise_on_non_zero_exit: false old_version = capture_with_info(*MRSK.app.current_running_version).strip.presence
execute *MRSK.app.start execute *MRSK.app.start
sleep 10
execute *MRSK.app.stop(version: old_version), raise_on_non_zero_exit: false
end end
audit_broadcast "Rolled back app to version #{version}" audit_broadcast "Rolled back app to version #{version}"

View File

@@ -18,8 +18,10 @@ class Mrsk::Commands::App < Mrsk::Commands::Base
docker :start, service_with_version docker :start, service_with_version
end end
def stop def stop(version: nil)
pipe current_container_id, xargs(docker(:stop)) pipe \
version ? container_id_for_version(version) : current_container_id,
xargs(docker(:stop))
end end
def info def info
@@ -132,6 +134,10 @@ class Mrsk::Commands::App < Mrsk::Commands::Base
end end
end end
def container_id_for_version(version)
container_id_for(container_name: service_with_version(version))
end
def service_filter def service_filter
[ "--filter", "label=service=#{config.service}" ] [ "--filter", "label=service=#{config.service}" ]
end end

View File

@@ -61,7 +61,7 @@ class Mrsk::Configuration::Role
"traefik.http.routers.#{config.service}.rule" => "PathPrefix(`/`)", "traefik.http.routers.#{config.service}.rule" => "PathPrefix(`/`)",
"traefik.http.services.#{config.service}.loadbalancer.healthcheck.path" => config.healthcheck["path"], "traefik.http.services.#{config.service}.loadbalancer.healthcheck.path" => config.healthcheck["path"],
"traefik.http.services.#{config.service}.loadbalancer.healthcheck.interval" => "1s", "traefik.http.services.#{config.service}.loadbalancer.healthcheck.interval" => "1s",
"traefik.http.middlewares.#{config.service}.retry.attempts" => "3", "traefik.http.middlewares.#{config.service}.retry.attempts" => "5",
"traefik.http.middlewares.#{config.service}.retry.initialinterval" => "500ms" "traefik.http.middlewares.#{config.service}.retry.initialinterval" => "500ms"
} }
else else

View File

@@ -2,7 +2,16 @@ require_relative "cli_test_case"
class CliAppTest < CliTestCase class CliAppTest < CliTestCase
test "boot" do test "boot" do
assert_match /Running docker run --detach --restart unless-stopped/, run_command("boot") # Stub current version fetch
SSHKit::Backend::Abstract.any_instance.stubs(:capture)
.returns("999") # new version
.then
.returns("123") # old version
run_command("boot").tap do |output|
assert_match /docker run --detach --restart unless-stopped/, output
assert_match /docker container ls --all --filter name=app-123 --quiet | xargs docker stop/, output
end
end end
test "boot will reboot if same version is already running" do test "boot will reboot if same version is already running" do
@@ -11,12 +20,17 @@ class CliAppTest < CliTestCase
# Prevent expected failures from outputting to terminal # Prevent expected failures from outputting to terminal
Thread.report_on_exception = false Thread.report_on_exception = false
MRSK.app.stubs(:run).raises(SSHKit::Command::Failed.new("already in use")).then.returns([ :docker, :run ]) MRSK.app.stubs(:run)
.raises(SSHKit::Command::Failed.new("already in use"))
.then
.raises(SSHKit::Command::Failed.new("already in use"))
.then
.returns([ :docker, :run ])
run_command("boot").tap do |output| run_command("boot").tap do |output|
assert_match /Rebooting container with same version already deployed/, output # Can't start what's already running assert_match /Rebooting container with same version 999 already deployed/, output # Can't start what's already running
assert_match /docker ps --quiet --filter label=service=app \| xargs docker stop/, output # Stop what's running assert_match /docker container ls --all --filter name=app-999 --quiet | xargs docker container rm/, output # Stop old running
assert_match /docker container ls --all --filter name=app-999 --quiet \| xargs docker container rm/, output # Remove old container assert_match /docker container ls --all --filter name=app-999 --quiet | xargs docker container rm/, output # Remove old container
assert_match /docker run/, output # Start new container assert_match /docker run/, output # Start new container
end end
ensure ensure

View File

@@ -14,7 +14,7 @@ class CommandsAppTest < ActiveSupport::TestCase
test "run" do test "run" do
assert_equal \ assert_equal \
"docker run --detach --restart unless-stopped --log-opt max-size=10m --name app-999 -e RAILS_MASTER_KEY=\"456\" --label service=\"app\" --label role=\"web\" --label traefik.http.routers.app.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.services.app.loadbalancer.healthcheck.path=\"/up\" --label traefik.http.services.app.loadbalancer.healthcheck.interval=\"1s\" --label traefik.http.middlewares.app.retry.attempts=\"3\" --label traefik.http.middlewares.app.retry.initialinterval=\"500ms\" dhh/app:999", "docker run --detach --restart unless-stopped --log-opt max-size=10m --name app-999 -e RAILS_MASTER_KEY=\"456\" --label service=\"app\" --label role=\"web\" --label traefik.http.routers.app.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.services.app.loadbalancer.healthcheck.path=\"/up\" --label traefik.http.services.app.loadbalancer.healthcheck.interval=\"1s\" --label traefik.http.middlewares.app.retry.attempts=\"5\" --label traefik.http.middlewares.app.retry.initialinterval=\"500ms\" dhh/app:999",
@app.run.join(" ") @app.run.join(" ")
end end
@@ -22,7 +22,7 @@ class CommandsAppTest < ActiveSupport::TestCase
@config[:volumes] = ["/local/path:/container/path" ] @config[:volumes] = ["/local/path:/container/path" ]
assert_equal \ assert_equal \
"docker run --detach --restart unless-stopped --log-opt max-size=10m --name app-999 -e RAILS_MASTER_KEY=\"456\" --volume /local/path:/container/path --label service=\"app\" --label role=\"web\" --label traefik.http.routers.app.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.services.app.loadbalancer.healthcheck.path=\"/up\" --label traefik.http.services.app.loadbalancer.healthcheck.interval=\"1s\" --label traefik.http.middlewares.app.retry.attempts=\"3\" --label traefik.http.middlewares.app.retry.initialinterval=\"500ms\" dhh/app:999", "docker run --detach --restart unless-stopped --log-opt max-size=10m --name app-999 -e RAILS_MASTER_KEY=\"456\" --volume /local/path:/container/path --label service=\"app\" --label role=\"web\" --label traefik.http.routers.app.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.services.app.loadbalancer.healthcheck.path=\"/up\" --label traefik.http.services.app.loadbalancer.healthcheck.interval=\"1s\" --label traefik.http.middlewares.app.retry.attempts=\"5\" --label traefik.http.middlewares.app.retry.initialinterval=\"500ms\" dhh/app:999",
@app.run.join(" ") @app.run.join(" ")
end end
@@ -30,7 +30,7 @@ class CommandsAppTest < ActiveSupport::TestCase
@config[:healthcheck] = { "path" => "/healthz" } @config[:healthcheck] = { "path" => "/healthz" }
assert_equal \ assert_equal \
"docker run --detach --restart unless-stopped --log-opt max-size=10m --name app-999 -e RAILS_MASTER_KEY=\"456\" --label service=\"app\" --label role=\"web\" --label traefik.http.routers.app.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.services.app.loadbalancer.healthcheck.path=\"/healthz\" --label traefik.http.services.app.loadbalancer.healthcheck.interval=\"1s\" --label traefik.http.middlewares.app.retry.attempts=\"3\" --label traefik.http.middlewares.app.retry.initialinterval=\"500ms\" dhh/app:999", "docker run --detach --restart unless-stopped --log-opt max-size=10m --name app-999 -e RAILS_MASTER_KEY=\"456\" --label service=\"app\" --label role=\"web\" --label traefik.http.routers.app.rule=\"PathPrefix(\\`/\\`)\" --label traefik.http.services.app.loadbalancer.healthcheck.path=\"/healthz\" --label traefik.http.services.app.loadbalancer.healthcheck.interval=\"1s\" --label traefik.http.middlewares.app.retry.attempts=\"5\" --label traefik.http.middlewares.app.retry.initialinterval=\"500ms\" dhh/app:999",
@app.run.join(" ") @app.run.join(" ")
end end

View File

@@ -42,7 +42,7 @@ class ConfigurationRoleTest < ActiveSupport::TestCase
end end
test "special label args for web" do test "special label args for web" do
assert_equal [ "--label", "service=\"app\"", "--label", "role=\"web\"", "--label", "traefik.http.routers.app.rule=\"PathPrefix(\\`/\\`)\"", "--label", "traefik.http.services.app.loadbalancer.healthcheck.path=\"/up\"", "--label", "traefik.http.services.app.loadbalancer.healthcheck.interval=\"1s\"", "--label", "traefik.http.middlewares.app.retry.attempts=\"3\"", "--label", "traefik.http.middlewares.app.retry.initialinterval=\"500ms\""], @config.role(:web).label_args assert_equal [ "--label", "service=\"app\"", "--label", "role=\"web\"", "--label", "traefik.http.routers.app.rule=\"PathPrefix(\\`/\\`)\"", "--label", "traefik.http.services.app.loadbalancer.healthcheck.path=\"/up\"", "--label", "traefik.http.services.app.loadbalancer.healthcheck.interval=\"1s\"", "--label", "traefik.http.middlewares.app.retry.attempts=\"5\"", "--label", "traefik.http.middlewares.app.retry.initialinterval=\"500ms\""], @config.role(:web).label_args
end end
test "custom labels" do test "custom labels" do
@@ -66,7 +66,7 @@ class ConfigurationRoleTest < ActiveSupport::TestCase
c[:servers]["beta"] = { "traefik" => "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", "traefik.http.routers.app.rule=\"PathPrefix(\\`/\\`)\"", "--label", "traefik.http.services.app.loadbalancer.healthcheck.path=\"/up\"", "--label", "traefik.http.services.app.loadbalancer.healthcheck.interval=\"1s\"", "--label", "traefik.http.middlewares.app.retry.attempts=\"3\"", "--label", "traefik.http.middlewares.app.retry.initialinterval=\"500ms\"" ], config.role(:beta).label_args assert_equal [ "--label", "service=\"app\"", "--label", "role=\"beta\"", "--label", "traefik.http.routers.app.rule=\"PathPrefix(\\`/\\`)\"", "--label", "traefik.http.services.app.loadbalancer.healthcheck.path=\"/up\"", "--label", "traefik.http.services.app.loadbalancer.healthcheck.interval=\"1s\"", "--label", "traefik.http.middlewares.app.retry.attempts=\"5\"", "--label", "traefik.http.middlewares.app.retry.initialinterval=\"500ms\"" ], config.role(:beta).label_args
end end
test "env overwritten by role" do test "env overwritten by role" do