Skip to content
Permalink
41 contributors

Users who have contributed to this file

@evanphx @nateberkopec @schneems @MSP-Greg @wjordan @Jesus @frankwong15 @chulkilee @ylecuyer @nathanallen @sthirugn @rafaelfranca
821 lines (746 sloc) 25.7 KB
# frozen_string_literal: true
require 'puma/const'
module Puma
# The methods that are available for use inside the configuration file.
# These same methods are used in Puma cli and the rack handler
# internally.
#
# Used manually (via CLI class):
#
# config = Configuration.new({}) do |user_config|
# user_config.port 3001
# end
# config.load
#
# puts config.options[:binds] # => "tcp://127.0.0.1:3001"
#
# Used to load file:
#
# $ cat puma_config.rb
# port 3002
#
# Resulting configuration:
#
# config = Configuration.new(config_file: "puma_config.rb")
# config.load
#
# puts config.options[:binds] # => "tcp://127.0.0.1:3002"
#
# You can also find many examples being used by the test suite in
# +test/config+.
#
class DSL
include ConfigDefault
def initialize(options, config)
@config = config
@options = options
@plugins = []
end
def _load_from(path)
if path
@path = path
instance_eval(File.read(path), path, 1)
end
ensure
_offer_plugins
end
def _offer_plugins
@plugins.each do |o|
if o.respond_to? :config
@options.shift
o.config self
end
end
@plugins.clear
end
def set_default_host(host)
@options[:default_host] = host
end
def default_host
@options[:default_host] || Configuration::DefaultTCPHost
end
def inject(&blk)
instance_eval(&blk)
end
def get(key,default=nil)
@options[key.to_sym] || default
end
# Load the named plugin for use by this configuration
#
def plugin(name)
@plugins << @config.load_plugin(name)
end
# Use an object or block as the rack application. This allows the
# configuration file to be the application itself.
#
# @example
# app do |env|
# body = 'Hello, World!'
#
# [
# 200,
# {
# 'Content-Type' => 'text/plain',
# 'Content-Length' => body.length.to_s
# },
# [body]
# ]
# end
#
# @see Puma::Configuration#app
#
def app(obj=nil, &block)
obj ||= block
raise "Provide either a #call'able or a block" unless obj
@options[:app] = obj
end
# Start the Puma control rack application on +url+. This application can
# be communicated with to control the main server. Additionally, you can
# provide an authentication token, so all requests to the control server
# will need to include that token as a query parameter. This allows for
# simple authentication.
#
# Check out {Puma::App::Status} to see what the app has available.
#
# @example
# activate_control_app 'unix:///var/run/pumactl.sock'
# @example
# activate_control_app 'unix:///var/run/pumactl.sock', { auth_token: '12345' }
# @example
# activate_control_app 'unix:///var/run/pumactl.sock', { no_token: true }
def activate_control_app(url="auto", opts={})
if url == "auto"
path = Configuration.temp_path
@options[:control_url] = "unix://#{path}"
@options[:control_url_temp] = path
else
@options[:control_url] = url
end
if opts[:no_token]
# We need to use 'none' rather than :none because this value will be
# passed on to an instance of OptionParser, which doesn't support
# symbols as option values.
#
# See: https://github.com/puma/puma/issues/1193#issuecomment-305995488
auth_token = 'none'
else
auth_token = opts[:auth_token]
auth_token ||= Configuration.random_token
end
@options[:control_auth_token] = auth_token
@options[:control_url_umask] = opts[:umask] if opts[:umask]
end
# Load additional configuration from a file
# Files get loaded later via Configuration#load
def load(file)
@options[:config_files] ||= []
@options[:config_files] << file
end
# Bind the server to +url+. "tcp://", "unix://" and "ssl://" are the only
# accepted protocols. Multiple urls can be bound to, calling `bind` does
# not overwrite previous bindings.
#
# The default is "tcp://0.0.0.0:9292".
#
# You can use query parameters within the url to specify options:
#
# * Set the socket backlog depth with +backlog+, default is 1024.
# * Set up an SSL certificate with +key+ & +cert+.
# * Set whether to optimize for low latency instead of throughput with
# +low_latency+, default is to optimize for low latency. This is done
# via +Socket::TCP_NODELAY+.
# * Set socket permissions with +umask+.
#
# @example Backlog depth
# bind 'unix:///var/run/puma.sock?backlog=512'
# @example SSL cert
# bind 'ssl://127.0.0.1:9292?key=key.key&cert=cert.pem'
# @example Disable optimization for low latency
# bind 'tcp://0.0.0.0:9292?low_latency=false'
# @example Socket permissions
# bind 'unix:///var/run/puma.sock?umask=0111'
# @see Puma::Runner#load_and_bind
# @see Puma::Cluster#run
#
def bind(url)
@options[:binds] ||= []
@options[:binds] << url
end
def clear_binds!
@options[:binds] = []
end
# Define the TCP port to bind to. Use +bind+ for more advanced options.
#
# @example
# port 9292
def port(port, host=nil)
host ||= default_host
bind "tcp://#{host}:#{port}"
end
# Define how long persistent connections can be idle before Puma closes them.
# @see Puma::Server.new
def persistent_timeout(seconds)
@options[:persistent_timeout] = Integer(seconds)
end
# Define how long the tcp socket stays open, if no data has been received.
# @see Puma::Server.new
def first_data_timeout(seconds)
@options[:first_data_timeout] = Integer(seconds)
end
# Work around leaky apps that leave garbage in Thread locals
# across requests.
def clean_thread_locals(which=true)
@options[:clean_thread_locals] = which
end
# When shutting down, drain the accept socket of pending connections and
# process them. This loops over the accept socket until there are no more
# read events and then stops looking and waits for the requests to finish.
# @see Puma::Server#graceful_shutdown
#
def drain_on_shutdown(which=true)
@options[:drain_on_shutdown] = which
end
# Set the environment in which the rack's app will run. The value must be
# a string.
#
# The default is "development".
#
# @example
# environment 'production'
def environment(environment)
@options[:environment] = environment
end
# How long to wait for threads to stop when shutting them
# down. Defaults to :forever. Specifying :immediately will cause
# Puma to kill the threads immediately. Otherwise the value
# is the number of seconds to wait.
#
# Puma always waits a few seconds after killing a thread for it to try
# to finish up it's work, even in :immediately mode.
# @see Puma::Server#graceful_shutdown
def force_shutdown_after(val=:forever)
i = case val
when :forever
-1
when :immediately
0
else
Float(val)
end
@options[:force_shutdown_after] = i
end
# Code to run before doing a restart. This code should
# close log files, database connections, etc.
#
# This can be called multiple times to add code each time.
#
# @example
# on_restart do
# puts 'On restart...'
# end
def on_restart(&block)
@options[:on_restart] ||= []
@options[:on_restart] << block
end
# Command to use to restart Puma. This should be just how to
# load Puma itself (ie. 'ruby -Ilib bin/puma'), not the arguments
# to Puma, as those are the same as the original process.
#
# @example
# restart_command '/u/app/lolcat/bin/restart_puma'
def restart_command(cmd)
@options[:restart_cmd] = cmd.to_s
end
# Store the pid of the server in the file at "path".
#
# @example
# pidfile '/u/apps/lolcat/tmp/pids/puma.pid'
def pidfile(path)
@options[:pidfile] = path.to_s
end
# Disable request logging, if this isn't used it'll be enabled by default.
#
# @example
# quiet
def quiet(which=true)
@options[:log_requests] = !which
end
# Enable request logging
#
def log_requests(which=true)
@options[:log_requests] = which
end
# Show debugging info
#
def debug
@options[:debug] = true
end
# Load +path+ as a rackup file.
#
# The default is "config.ru".
#
# @example
# rackup '/u/apps/lolcat/config.ru'
def rackup(path)
@options[:rackup] ||= path.to_s
end
def early_hints(answer=true)
@options[:early_hints] = answer
end
# Redirect +STDOUT+ and +STDERR+ to files specified. The +append+ parameter
# specifies whether the output is appended, the default is +false+.
#
# @example
# stdout_redirect '/app/lolcat/log/stdout', '/app/lolcat/log/stderr'
# @example
# stdout_redirect '/app/lolcat/log/stdout', '/app/lolcat/log/stderr', true
def stdout_redirect(stdout=nil, stderr=nil, append=false)
@options[:redirect_stdout] = stdout
@options[:redirect_stderr] = stderr
@options[:redirect_append] = append
end
def log_formatter(&block)
@options[:log_formatter] = block
end
# Configure +min+ to be the minimum number of threads to use to answer
# requests and +max+ the maximum.
#
# The default is "0, 16".
#
# @example
# threads 0, 16
# @example
# threads 5, 5
def threads(min, max)
min = Integer(min)
max = Integer(max)
if min > max
raise "The minimum (#{min}) number of threads must be less than or equal to the max (#{max})"
end
if max < 1
raise "The maximum number of threads (#{max}) must be greater than 0"
end
@options[:min_threads] = min
@options[:max_threads] = max
end
# Instead of `bind 'ssl://127.0.0.1:9292?key=key_path&cert=cert_path'` you
# can also use the this method.
#
# @example
# ssl_bind '127.0.0.1', '9292', {
# cert: path_to_cert,
# key: path_to_key,
# ssl_cipher_filter: cipher_filter, # optional
# verify_mode: verify_mode, # default 'none'
# }
# @example For JRuby additional keys are required: keystore & keystore_pass.
# ssl_bind '127.0.0.1', '9292', {
# cert: path_to_cert,
# key: path_to_key,
# ssl_cipher_filter: cipher_filter, # optional
# verify_mode: verify_mode, # default 'none'
# keystore: path_to_keystore,
# keystore_pass: password
# }
def ssl_bind(host, port, opts)
verify = opts.fetch(:verify_mode, 'none').to_s
no_tlsv1 = opts.fetch(:no_tlsv1, 'false')
no_tlsv1_1 = opts.fetch(:no_tlsv1_1, 'false')
ca_additions = "&ca=#{opts[:ca]}" if ['peer', 'force_peer'].include?(verify)
if defined?(JRUBY_VERSION)
keystore_additions = "keystore=#{opts[:keystore]}&keystore-pass=#{opts[:keystore_pass]}"
bind "ssl://#{host}:#{port}?cert=#{opts[:cert]}&key=#{opts[:key]}&#{keystore_additions}&verify_mode=#{verify}&no_tlsv1=#{no_tlsv1}&no_tlsv1_1=#{no_tlsv1_1}#{ca_additions}"
else
ssl_cipher_filter = "&ssl_cipher_filter=#{opts[:ssl_cipher_filter]}" if opts[:ssl_cipher_filter]
bind "ssl://#{host}:#{port}?cert=#{opts[:cert]}&key=#{opts[:key]}#{ssl_cipher_filter}&verify_mode=#{verify}&no_tlsv1=#{no_tlsv1}&no_tlsv1_1=#{no_tlsv1_1}#{ca_additions}"
end
end
# Use +path+ as the file to store the server info state. This is
# used by +pumactl+ to query and control the server.
#
# @example
# state_path '/u/apps/lolcat/tmp/pids/puma.state'
def state_path(path)
@options[:state] = path.to_s
end
# Use +permission+ to restrict permissions for the state file.
#
# @example
# state_permission 0600
# @version 5.0.0
#
def state_permission(permission)
@options[:state_permission] = permission
end
# How many worker processes to run. Typically this is set to
# the number of available cores.
#
# The default is 0.
#
# @note Cluster mode only.
# @see Puma::Cluster
def workers(count)
@options[:workers] = count.to_i
end
# Code to run immediately before master process
# forks workers (once on boot). These hooks can block if necessary
# to wait for background operations unknown to Puma to finish before
# the process terminates.
# This can be used to close any connections to remote servers (database,
# Redis, ...) that were opened when preloading the code.
#
# This can be called multiple times to add several hooks.
#
# @note Cluster mode only.
# @example
# before_fork do
# puts "Starting workers..."
# end
def before_fork(&block)
@options[:before_fork] ||= []
@options[:before_fork] << block
end
# Code to run in a worker when it boots to setup
# the process before booting the app.
#
# This can be called multiple times to add several hooks.
#
# @note Cluster mode only.
# @example
# on_worker_boot do
# puts 'Before worker boot...'
# end
def on_worker_boot(&block)
@options[:before_worker_boot] ||= []
@options[:before_worker_boot] << block
end
# Code to run immediately before a worker shuts
# down (after it has finished processing HTTP requests). These hooks
# can block if necessary to wait for background operations unknown
# to Puma to finish before the process terminates.
#
# This can be called multiple times to add several hooks.
#
# @note Cluster mode only.
# @example
# on_worker_shutdown do
# puts 'On worker shutdown...'
# end
def on_worker_shutdown(&block)
@options[:before_worker_shutdown] ||= []
@options[:before_worker_shutdown] << block
end
# Code to run in the master right before a worker is started. The worker's
# index is passed as an argument.
#
# This can be called multiple times to add several hooks.
#
# @note Cluster mode only.
# @example
# on_worker_fork do
# puts 'Before worker fork...'
# end
def on_worker_fork(&block)
@options[:before_worker_fork] ||= []
@options[:before_worker_fork] << block
end
# Code to run in the master after a worker has been started. The worker's
# index is passed as an argument.
#
# This is called everytime a worker is to be started.
#
# @note Cluster mode only.
# @example
# after_worker_fork do
# puts 'After worker fork...'
# end
def after_worker_fork(&block)
@options[:after_worker_fork] ||= []
@options[:after_worker_fork] = block
end
alias_method :after_worker_boot, :after_worker_fork
# When `fork_worker` is enabled, code to run in Worker 0
# before all other workers are re-forked from this process,
# after the server has temporarily stopped serving requests
# (once per complete refork cycle).
#
# This can be used to trigger extra garbage-collection to maximize
# copy-on-write efficiency, or close any connections to remote servers
# (database, Redis, ...) that were opened while the server was running.
#
# This can be called multiple times to add several hooks.
#
# @note Cluster mode with `fork_worker` enabled only.
# @example
# on_refork do
# 3.times {GC.start}
# end
# @version 5.0.0
#
def on_refork(&block)
@options[:before_refork] ||= []
@options[:before_refork] << block
end
# Code to run out-of-band when the worker is idle.
# These hooks run immediately after a request has finished
# processing and there are no busy threads on the worker.
# The worker doesn't accept new requests until this code finishes.
#
# This hook is useful for running out-of-band garbage collection
# or scheduling asynchronous tasks to execute after a response.
#
# This can be called multiple times to add several hooks.
def out_of_band(&block)
@options[:out_of_band] ||= []
@options[:out_of_band] << block
end
# The directory to operate out of.
#
# The default is the current directory.
#
# @example
# directory '/u/apps/lolcat'
def directory(dir)
@options[:directory] = dir.to_s
end
# Preload the application before starting the workers; this conflicts with
# phased restart feature. This is off by default.
#
# @note Cluster mode only.
# @example
# preload_app!
def preload_app!(answer=true)
@options[:preload_app] = answer
end
# Use +obj+ or +block+ as the low level error handler. This allows the
# configuration file to change the default error on the server.
#
# @example
# lowlevel_error_handler do |err|
# [200, {}, ["error page"]]
# end
def lowlevel_error_handler(obj=nil, &block)
obj ||= block
raise "Provide either a #call'able or a block" unless obj
@options[:lowlevel_error_handler] = obj
end
# This option is used to allow your app and its gems to be
# properly reloaded when not using preload.
#
# When set, if Puma detects that it's been invoked in the
# context of Bundler, it will cleanup the environment and
# re-run itself outside the Bundler environment, but directly
# using the files that Bundler has setup.
#
# This means that Puma is now decoupled from your Bundler
# context and when each worker loads, it will be loading a
# new Bundler context and thus can float around as the release
# dictates.
#
# @see extra_runtime_dependencies
#
# @note This is incompatible with +preload_app!+.
# @note This is only supported for RubyGems 2.2+
def prune_bundler(answer=true)
@options[:prune_bundler] = answer
end
# By default, Puma will raise SignalException when SIGTERM is received. In
# environments where SIGTERM is something expected, you can suppress these
# with this option.
#
# This can be useful for example in Kubernetes, where rolling restart is
# guaranteed usually on infrastructure level.
#
# @example
# raise_exception_on_sigterm false
# @see Puma::Launcher#setup_signals
# @see Puma::Cluster#setup_signals
#
def raise_exception_on_sigterm(answer=true)
@options[:raise_exception_on_sigterm] = answer
end
# When using prune_bundler, if extra runtime dependencies need to be loaded to
# initialize your app, then this setting can be used. This includes any Puma plugins.
#
# Before bundler is pruned, the gem names supplied will be looked up in the bundler
# context and then loaded again after bundler is pruned.
# Only applies if prune_bundler is used.
#
# @example
# extra_runtime_dependencies ['gem_name_1', 'gem_name_2']
# @example
# extra_runtime_dependencies ['puma_worker_killer', 'puma-heroku']
# @see Puma::Launcher#extra_runtime_deps_directories
#
def extra_runtime_dependencies(answer = [])
@options[:extra_runtime_dependencies] = Array(answer)
end
# Additional text to display in process listing.
#
# If you do not specify a tag, Puma will infer it. If you do not want Puma
# to add a tag, use an empty string.
#
# @example
# tag 'app name'
# @example
# tag ''
def tag(string)
@options[:tag] = string.to_s
end
# Verifies that all workers have checked in to the master process within
# the given timeout. If not the worker process will be restarted. This is
# not a request timeout, it is to protect against a hung or dead process.
# Setting this value will not protect against slow requests.
#
# The minimum value is 6 seconds, the default value is 60 seconds.
#
# @note Cluster mode only.
# @example
# worker_timeout 60
# @see Puma::Cluster::Worker#ping_timeout
#
def worker_timeout(timeout)
timeout = Integer(timeout)
min = Const::WORKER_CHECK_INTERVAL
if timeout <= min
raise "The minimum worker_timeout must be greater than the worker reporting interval (#{min})"
end
@options[:worker_timeout] = timeout
end
# Change the default worker timeout for booting.
#
# If unspecified, this defaults to the value of worker_timeout.
#
# @note Cluster mode only.
#
# @example
# worker_boot_timeout 60
# @see Puma::Cluster::Worker#ping_timeout
#
def worker_boot_timeout(timeout)
@options[:worker_boot_timeout] = Integer(timeout)
end
# Set the timeout for worker shutdown.
#
# @note Cluster mode only.
# @see Puma::Cluster::Worker#term
#
def worker_shutdown_timeout(timeout)
@options[:worker_shutdown_timeout] = Integer(timeout)
end
# When set to true (the default), workers accept all requests
# and queue them before passing them to the handlers.
# When set to false, each worker process accepts exactly as
# many requests as it is configured to simultaneously handle.
#
# Queueing requests generally improves performance. In some
# cases, such as a single threaded application, it may be
# better to ensure requests get balanced across workers.
#
# Note that setting this to false disables HTTP keepalive and
# slow clients will occupy a handler thread while the request
# is being sent. A reverse proxy, such as nginx, can handle
# slow clients and queue requests before they reach Puma.
# @see Puma::Server
def queue_requests(answer=true)
@options[:queue_requests] = answer
end
# When a shutdown is requested, the backtraces of all the
# threads will be written to $stdout. This can help figure
# out why shutdown is hanging.
#
def shutdown_debug(val=true)
@options[:shutdown_debug] = val
end
# Attempts to route traffic to less-busy workers by causing them to delay
# listening on the socket, allowing workers which are not processing any
# requests to pick up new requests first.
#
# Only works on MRI. For all other interpreters, this setting does nothing.
# @see Puma::Server#handle_servers
# @see Puma::ThreadPool#wait_for_less_busy_worker
# @version 5.0.0
#
def wait_for_less_busy_worker(val=0.005)
@options[:wait_for_less_busy_worker] = val.to_f
end
# Control how the remote address of the connection is set. This
# is configurable because to calculate the true socket peer address
# a kernel syscall is required which for very fast rack handlers
# slows down the handling significantly.
#
# There are 4 possible values:
#
# 1. **:socket** (the default) - read the peername from the socket using the
# syscall. This is the normal behavior.
# 2. **:localhost** - set the remote address to "127.0.0.1"
# 3. **header: <http_header>**- set the remote address to the value of the
# provided http header. For instance:
# `set_remote_address header: "X-Real-IP"`.
# Only the first word (as separated by spaces or comma) is used, allowing
# headers such as X-Forwarded-For to be used as well.
# 4. **\<Any string\>** - this allows you to hardcode remote address to any value
# you wish. Because Puma never uses this field anyway, it's format is
# entirely in your hands.
#
def set_remote_address(val=:socket)
case val
when :socket
@options[:remote_address] = val
when :localhost
@options[:remote_address] = :value
@options[:remote_address_value] = "127.0.0.1".freeze
when String
@options[:remote_address] = :value
@options[:remote_address_value] = val
when Hash
if hdr = val[:header]
@options[:remote_address] = :header
@options[:remote_address_header] = "HTTP_" + hdr.upcase.tr("-", "_")
else
raise "Invalid value for set_remote_address - #{val.inspect}"
end
else
raise "Invalid value for set_remote_address - #{val}"
end
end
# When enabled, workers will be forked from worker 0 instead of from the master process.
# This option is similar to `preload_app` because the app is preloaded before forking,
# but it is compatible with phased restart.
#
# This option also enables the `refork` command (SIGURG), which optimizes copy-on-write performance
# in a running app.
#
# A refork will automatically trigger once after the specified number of requests
# (default 1000), or pass 0 to disable auto refork.
#
# @note Cluster mode only.
# @version 5.0.0
#
def fork_worker(after_requests=1000)
@options[:fork_worker] = Integer(after_requests)
end
# When enabled, Puma will GC 4 times before forking workers.
# If available (Ruby 2.7+), we will also call GC.compact.
# Not recommended for non-MRI Rubies.
#
# Based on the work of Koichi Sasada and Aaron Patterson, this option may
# decrease memory utilization of preload-enabled cluster-mode Pumas. It will
# also increase time to boot and fork. See your logs for details on how much
# time this adds to your boot process. For most apps, it will be less than one
# second.
#
# @see Puma::Cluster#nakayoshi_gc
# @version 5.0.0
#
def nakayoshi_fork(enabled=true)
@options[:nakayoshi_fork] = enabled
end
# The number of requests to attempt inline before sending a client back to
# the reactor to be subject to normal ordering.
#
def max_fast_inline(num_of_requests)
@options[:max_fast_inline] = Float(num_of_requests)
end
end
end
You can’t perform that action at this time.