Zero downtime deployment with cord file
When replacing a container currently we:
1. Boot the new container
2. Wait for it to become healthy
3. Stop the old container
Traefik will send requests to the old container until it notices that it
is unhealthy. But it may have stopped serving requests before that point
which can result in errors.
To get round that the new boot process is:
1. Create a directory with a single file on the host
2. Boot the new container, mounting the cord file into /tmp and
including a check for the file in the docker healthcheck
3. Wait for it to become healthy
4. Delete the healthcheck file ("cut the cord") for the old container
5. Wait for it to become unhealthy and give Traefik a couple of seconds
to notice
6. Stop the old container
The extra steps ensure that Traefik stops sending requests before the
old container is shutdown.
This commit is contained in:
@@ -1,4 +1,5 @@
|
||||
class Kamal::Configuration::Role
|
||||
CORD_FILE = "cord"
|
||||
delegate :argumentize, :env_file_with_secrets, :optionize, to: Kamal::Utils
|
||||
|
||||
attr_accessor :name
|
||||
@@ -47,28 +48,52 @@ class Kamal::Configuration::Role
|
||||
argumentize "--env-file", host_env_file_path
|
||||
end
|
||||
|
||||
def health_check_args
|
||||
def health_check_args(cord: true)
|
||||
if health_check_cmd.present?
|
||||
optionize({ "health-cmd" => health_check_cmd, "health-interval" => health_check_interval })
|
||||
if cord && uses_cord?
|
||||
optionize({ "health-cmd" => health_check_cmd_with_cord, "health-interval" => health_check_interval })
|
||||
.concat(["--volume", "#{cord_host_directory}:#{cord_container_directory}"])
|
||||
else
|
||||
optionize({ "health-cmd" => health_check_cmd, "health-interval" => health_check_interval })
|
||||
end
|
||||
else
|
||||
[]
|
||||
end
|
||||
end
|
||||
|
||||
def health_check_cmd
|
||||
options = specializations["healthcheck"] || {}
|
||||
options = config.healthcheck.merge(options) if running_traefik?
|
||||
health_check_options["cmd"] || http_health_check(port: health_check_options["port"], path: health_check_options["path"])
|
||||
end
|
||||
|
||||
options["cmd"] || http_health_check(port: options["port"], path: options["path"])
|
||||
def health_check_cmd_with_cord
|
||||
"(#{health_check_cmd}) && (stat #{cord_container_file} > /dev/null || exit 1)"
|
||||
end
|
||||
|
||||
def health_check_interval
|
||||
options = specializations["healthcheck"] || {}
|
||||
options = config.healthcheck.merge(options) if running_traefik?
|
||||
|
||||
options["interval"] || "1s"
|
||||
health_check_options["interval"] || "1s"
|
||||
end
|
||||
|
||||
def uses_cord?
|
||||
running_traefik? && cord_container_directory.present? && health_check_cmd.present?
|
||||
end
|
||||
|
||||
def cord_host_directory
|
||||
File.join config.run_directory_as_docker_volume, "cords", [full_name, config.run_id].join("-")
|
||||
end
|
||||
|
||||
def cord_host_file
|
||||
File.join cord_host_directory, CORD_FILE
|
||||
end
|
||||
|
||||
def cord_container_directory
|
||||
health_check_options.fetch("cord", nil)
|
||||
end
|
||||
|
||||
def cord_container_file
|
||||
File.join cord_container_directory, CORD_FILE
|
||||
end
|
||||
|
||||
|
||||
def cmd
|
||||
specializations["cmd"]
|
||||
end
|
||||
@@ -85,6 +110,10 @@ class Kamal::Configuration::Role
|
||||
name.web? || specializations["traefik"]
|
||||
end
|
||||
|
||||
def full_name
|
||||
[ config.service, name, config.destination ].compact.join("-")
|
||||
end
|
||||
|
||||
private
|
||||
attr_accessor :config
|
||||
|
||||
@@ -164,4 +193,12 @@ class Kamal::Configuration::Role
|
||||
def http_health_check(port:, path:)
|
||||
"curl -f #{URI.join("http://localhost:#{port}", path)} || exit 1" if path.present? || port.present?
|
||||
end
|
||||
|
||||
def health_check_options
|
||||
@health_check_options ||= begin
|
||||
options = specializations["healthcheck"] || {}
|
||||
options = config.healthcheck.merge(options) if running_traefik?
|
||||
options
|
||||
end
|
||||
end
|
||||
end
|
||||
|
||||
Reference in New Issue
Block a user