rails, puma auto-tune

#!/usr/bin/env ruby
# Instantiate about one process per X MiB of available memory, scaling up to as
# close to MAX_THREADS as possible while observing an upper bound based on the
# number of virtual/logical CPUs. If there are fewer processes than
# MAX_THREADS, add threads per process to reach MAX_THREADS.
require 'etc'

KB_PER_WORKER = 64 * 1_024 # average of peak PSS of single-threaded processes (watch smem -k)
MIN_WORKERS = 2
MAX_WORKERS_PER_VCPU = 1.25 # virtual/logical
MIN_THREADS_PER_WORKER = 1
MAX_THREADS = Integer(ENV['MAX_CONCURRENCY'] || 256)

def meminfo(arg)
  File.open('/proc/meminfo') do |f|
    f.each_line do |line|
      key, value = line.split(/:\s+/)
      return value.split(/\s+/).first.to_i if key == arg
    end
  end

  fail "Unable to find `#{arg}' in /proc/meminfo!"
end

def auto_tune
  avail_mem = meminfo('MemAvailable') * 0.8 - MAX_THREADS * 1_024

  workers = [
    [(1.0 * avail_mem / KB_PER_WORKER).floor, MIN_WORKERS].max,
    (Etc.nprocessors * MAX_WORKERS_PER_VCPU).ceil
  ].min

  threads_per_worker = [
    workers < MAX_THREADS ? (1.0 * MAX_THREADS / workers).ceil : -Float::INFINITY,
    MIN_THREADS_PER_WORKER
  ].max

  [workers, threads_per_worker]
end

p auto_tune if $0 == __FILE__

 Puma can serve each request in a thread from an internal thread pool.
# The `threads` method setting takes two numbers a minimum and maximum.
# Any libraries that use thread pools should be configured to match
# the maximum value specified for Puma. Default is set to 5 threads for minimum
# and maximum, this matches the default thread size of Active Record.
#
threads_count = ENV.fetch("RAILS_MAX_THREADS") { 5 }.to_i
threads threads_count, threads_count

# Specifies the `port` that Puma will listen on to receive requests, default is 3000.
#
port        ENV.fetch("PORT") { 3000 }

# Specifies the `environment` that Puma will run in.
#
environment ENV.fetch("RAILS_ENV") { "development" }

# Specifies the number of `workers` to boot in clustered mode.
# Workers are forked webserver processes. If using threads and workers together
# the concurrency of the application would be max `threads` * `workers`.
# Workers do not work on JRuby or Windows (both of which do not support
# processes).
#
workers ENV.fetch("WEB_CONCURRENCY") { 2 }

# Use the `preload_app!` method when specifying a `workers` number.
# This directive tells Puma to first boot the application and load code
# before forking the application. This takes advantage of Copy On Write
# process behavior so workers use less memory. If you use this option
# you need to make sure to reconnect any threads in the `on_worker_boot`
# block.
#

preload_app!

# The code in the `on_worker_boot` will be called if you are using
# clustered mode by specifying a number of `workers`. After each worker
# process is booted this block will be run, if you are using `preload_app!`
# option you will want to use this block to reconnect to any threads
# or connections that may have been created at application boot, Ruby
# cannot share connections between processes.
#
# on_worker_boot do
#   ActiveRecord::Base.establish_connection if defined?(ActiveRecord)
# end

# Allow puma to be restarted by `rails restart` command.
plugin :tmp_restart