puma_worker_killer 0.0.6 → 0.0.7

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 948c271c454d3a1dc1b2bccaf270bcba98de73f8
-  data.tar.gz: 14b5ed3e5b5c28b03df31e23dc7a1047d1965a30
+  metadata.gz: 20389b0e2ba67ab795071554a0e52b6599cd3612
+  data.tar.gz: 0789e6a76e4c12394df935bd9b16b503d6990a62
 SHA512:
-  metadata.gz: d47c3c52bca9383936074705a88ef19e62751c183b0f086a1f004970655e5bf3b9bc934a2706013a4af3a69b160ad986371dfa36105b52acee0d6a981481f6a1
-  data.tar.gz: 683b00d578d35bc61b3450acbfb2276f438208b99f50e7951650c2348d2bdc0fc2570a5bf4f52b446131e80dc61ff21d39821c87c1cd6fe3e2768ff20cd2e58a
+  metadata.gz: 62aa47f0f5032df0d0887e8001143059b4107fe1871932b545d4e56820c0012260ea73ffb77aed9be3fe83a1a3ec08b2f23633a05f96d8375885b3cdf1e05454
+  data.tar.gz: 30fedd4919738c70432d7de4861089c03212c651805d4f564cd83e16d46f3b1d9bfc60ce44ae03a71cd1dff15b7547fbfb757b6b26d5e894f577ab0765669cad

data/.travis.yml

@@ -7,6 +7,8 @@ rvm:
   - 2.3.0
   - ruby-head
   - rbx
+before_install:
+  - gem install bundler -v 1.12.5
 
 matrix:
   allow_failures:

data/CHANGELOG.md

@@ -1,3 +1,7 @@
+## 0.0.7
+
+- Logging is configurable #41
+
 ## 0.0.6
 
 - Log PID of worker insead of inspecting the worker #33

data/README.md

@@ -29,15 +29,68 @@ Then run `$ bundle install`
 
 -->
 
-Somewhere in your main process run this code:
+## Turn on Rolling Restarts
+
+A rolling restart will kill each of your workers on a rolling basis. You set the frequency which it conducts the restart. This is a simple way to keep memory down as Ruby web programs generally increase memory usage over time. If you're using Heroku [it is difficult to measure RAM from inside of a container accurately](https://github.com/schneems/get_process_mem/issues/7), so it is recommended to use this feature or use a [log-drain-based worker killer](https://github.com/arches/whacamole). You can enable roling restarts by running:
 
 ```ruby
-# config/initializers/puma_worker_killer.rb
-PumaWorkerKiller.start
+before_fork do
+  require 'puma_worker_killer'
+
+  PumaWorkerKiller.enable_rolling_restart # Default is every 6 hours
+end
+
+```
+
+or you can pass in the restart frequency:
+
+```ruby
+PumaWorkerKiller.enable_rolling_restart(12 * 3600) # 12 hours in seconds
+```
+
+Make sure if you do this to not accidentally call `PumaWorkerKiller.start` as well.
+
+## Enable Worker Killing
+
+If you're not running on a containerized platform you can try to detect the amount of memory you're using and only kill Puma workers when you're over that limit. It may allow you to go for longer periods of time without killing a worker however it is more error prone than rolling restarts. To enable measurement based worker killing put this in your `config/puma.rb`:
+
+```ruby
+# config/puma.rb
+
+before_fork do
+  require 'puma_worker_killer'
+
+  PumaWorkerKiller.start
+end
 ```
 
 That's it. Now on a regular basis the size of all Puma and all of it's forked processes will be evaluated and if they're over the RAM threshold will be killed. Don't worry Puma will notice a process is missing a spawn a fresh copy with a much smaller RAM footprint ASAP.
 
+## Troubleshooting
+
+When you boot your program locally you should see debug output:
+
+```
+[77773] Puma starting in cluster mode...
+[77773] * Version 3.1.0 (ruby 2.3.1-p112), codename: El Niño Winter Wonderland
+[77773] * Min threads: 0, max threads: 16
+[77773] * Environment: development
+[77773] * Process workers: 2
+[77773] * Phased restart available
+[77773] * Listening on tcp://0.0.0.0:9292
+[77773] Use Ctrl-C to stop
+[77773] PumaWorkerKiller: Consuming 54.34765625 mb with master and 2 workers.
+```
+
+If you don't see any `PumaWorkerKiller` output, make sure that you are running with multiple workers. PWK only functions if you have workers enabled, you should see something like this when Puma boots:
+
+```
+[77773] * Process workers: 2
+```
+
+If you've configured PWK's frequency try reducing it to a very low value
+
+
 ## Configure
 
 Before calling `start` you can configure `PumaWorkerKiller`. You can do so using a configure block or calling methods directly:
@@ -48,13 +101,17 @@ PumaWorkerKiller.config do |config|
   config.frequency     = 5    # seconds
   config.percent_usage = 0.98
   config.rolling_restart_frequency = 12 * 3600 # 12 hours in seconds
+  config.reaper_status_logs = true # setting this to false will not log lines like:
+  # PumaWorkerKiller: Consuming 54.34765625 mb with master and 2 workers.
 end
 PumaWorkerKiller.start
 ```
 
 ## Attention
-If you start puma as a daemon, to add puma worker killer config into puma config file, rather than into initializers:    
-Sample like this: (in puma.rb file)
+
+If you start puma as a daemon, to add puma worker killer config into puma config file, rather than into initializers:
+Sample like this: (in `config/puma.rb` file):
+
 ```ruby
 before_fork do
   PumaWorkerKiller.config do |config|
@@ -93,27 +150,18 @@ PumaWorkerKiller.rolling_restart_frequency = 12 * 3600 # 12 hours in seconds
 
 By default PumaWorkerKiller will perform a rolling restart of all your worker processes every 12 hours. To disable, set to `false`.
 
-## Only turn on Rolling Restarts
-
-If you're running on a platform like [Heroku where it is difficult to measure RAM from inside of a container accurately](https://github.com/schneems/get_process_mem/issues/7), you may want to disable the "worker killer" functionality and only use the rolling restart. You can do that by running:
+You may want to hide the following log lines: `PumaWorkerKiller: Consuming 54.34765625 mb with master and 2 workers.`. To do that set:
 
 ```ruby
-PumaWorkerKiller.enable_rolling_restart # Default is every 6 hours
+PumaWorkerKiller.reaper_status_logs = false
 ```
 
-or you can pass in the restart frequency
-
-```ruby
-PumaWorkerKiller.enable_rolling_restart(12 * 3600) # 12 hours in seconds
-```
-
-Make sure if you do this to not accidentally call `PumaWorkerKiller.start` as well.
+Note: It is `true` by default.
 
 ## License
 
 MIT
 
-
 ## Feedback
 
 Open up an issue or ping me on twitter [@schneems](http://twitter.com/schneems).

data/lib/puma_worker_killer/auto_reap.rb

@@ -11,8 +11,8 @@ module PumaWorkerKiller
 
       Thread.new do
         while @running
-          @reaper.reap
           sleep @timeout
+          @reaper.reap
         end
       end
     end

data/lib/puma_worker_killer/puma_memory.rb

@@ -18,6 +18,10 @@ module PumaWorkerKiller
     # rescue Errno::ECHILD
     end
 
+    def workers_stopped?
+      !running?
+    end
+
     def running?
       @master && workers.any?
     end

data/lib/puma_worker_killer/reaper.rb

@@ -1,8 +1,9 @@
 module PumaWorkerKiller
   class Reaper
-    def initialize(max_ram, master = nil)
+    def initialize(max_ram, master = nil, reaper_status_logs = true)
       @cluster = PumaWorkerKiller::PumaMemory.new(master)
       @max_ram = max_ram
+      @reaper_status_logs = reaper_status_logs
     end
 
     # used for tes
@@ -11,11 +12,11 @@ module PumaWorkerKiller
     end
 
     def reap
-      return false unless @cluster.running?
+      return false if @cluster.workers_stopped?
       if (total = get_total_memory) > @max_ram
         @cluster.master.log "PumaWorkerKiller: Out of memory. #{@cluster.workers.count} workers consuming total: #{total} mb out of max: #{@max_ram} mb. Sending TERM to pid #{@cluster.largest_worker.pid} consuming #{@cluster.largest_worker_memory} mb."
         @cluster.term_largest_worker
-      else
+      elsif @reaper_status_logs
         @cluster.master.log "PumaWorkerKiller: Consuming #{total} mb with master and #{@cluster.workers.count} workers."
       end
     end

data/lib/puma_worker_killer/version.rb

@@ -1,3 +1,3 @@
 module PumaWorkerKiller
-  VERSION = "0.0.6"
+  VERSION = "0.0.7"
 end

data/lib/puma_worker_killer.rb

@@ -3,18 +3,19 @@ require 'get_process_mem'
 module PumaWorkerKiller
   extend self
 
-  attr_accessor :ram, :frequency, :percent_usage, :rolling_restart_frequency
+  attr_accessor :ram, :frequency, :percent_usage, :rolling_restart_frequency, :reaper_status_logs
   self.ram           = 512  # mb
   self.frequency     = 10   # seconds
   self.percent_usage = 0.99 # percent of RAM to use
   self.rolling_restart_frequency = 6 * 3600 # 6 hours in seconds
+  self.reaper_status_logs = true
 
   def config
     yield self
   end
 
-  def reaper(ram = self.ram, percent = self.percent_usage)
-    Reaper.new(ram * percent_usage)
+  def reaper(ram = self.ram, percent = self.percent_usage, reaper_status_logs = self.reaper_status_logs)
+    Reaper.new(ram * percent_usage, nil, reaper_status_logs)
   end
 
   def start(frequency = self.frequency, reaper = self.reaper)

data/puma_worker_killer.gemspec

@@ -21,6 +21,7 @@ Gem::Specification.new do |gem|
   gem.add_dependency "puma",              ">= 2.7", "< 4"
   gem.add_dependency "get_process_mem",   "~>  0.2"
   gem.add_development_dependency "rack", "~> 1.6"
+  gem.add_development_dependency "wait_for_it", "~> 0.1"
   gem.add_development_dependency "rake",  "~> 10.1"
   gem.add_development_dependency "test-unit",  ">= 0"
 

data/test/fixtures/big.ru

@@ -0,0 +1,10 @@
+load File.expand_path("../fixture_helper.rb", __FILE__)
+
+PumaWorkerKiller.start
+
+@memory = []
+10_000.times.each do
+  @memory << SecureRandom.hex
+end
+
+run HelloWorldApp

data/test/fixtures/config/puma_worker_killer_start.rb

@@ -0,0 +1,6 @@
+load File.expand_path("../../fixture_helper.rb", __FILE__)
+
+before_fork do
+  require 'puma_worker_killer'
+  PumaWorkerKiller.start
+end

data/test/fixtures/default.ru

@@ -0,0 +1,5 @@
+load File.expand_path("../fixture_helper.rb", __FILE__)
+
+PumaWorkerKiller.start
+
+run HelloWorldApp

data/test/fixtures/{app.ru → fixture_helper.rb}

@@ -1,3 +1,5 @@
+require 'securerandom'
+
 require 'rack'
 require 'rack/server'
 
@@ -7,21 +9,17 @@ PumaWorkerKiller.config do |config|
   config.ram       = Integer(ENV['PUMA_RAM'])       if ENV['PUMA_RAM']
   config.frequency = Integer(ENV['PUMA_FREQUENCY']) if ENV['PUMA_FREQUENCY']
 end
-PumaWorkerKiller.start
-
 
-puts "Frequency: #{PumaWorkerKiller.frequency}" if ENV['PUMA_FREQUENCY']
+puts "Frequency: #{ PumaWorkerKiller.frequency }" if ENV['PUMA_FREQUENCY']
 
 class HelloWorld
-  def response
+  def response(env)
     [200, {}, ['Hello World']]
   end
 end
 
 class HelloWorldApp
   def self.call(env)
-    HelloWorld.new.response
+    HelloWorld.new.response(env)
   end
 end
-
-run HelloWorldApp

data/test/fixtures/rolling_restart.ru

@@ -0,0 +1,5 @@
+load File.expand_path("../fixture_helper.rb", __FILE__)
+
+PumaWorkerKiller.enable_rolling_restart(1) # 1 second
+
+run HelloWorldApp

data/test/puma_worker_killer_test.rb

@@ -3,75 +3,51 @@ require 'test_helper'
 class PumaWorkerKillerTest < Test::Unit::TestCase
 
   def test_starts
-    app_path = fixture_path.join("app.ru")
     port     = 0 # http://stackoverflow.com/questions/200484/how-do-you-find-a-free-tcp-server-port-using-ruby
-    puma_log = Pathname.new "#{ SecureRandom.hex }-puma.log"
-    pid  = Process.spawn("PUMA_FREQUENCY=1 bundle exec puma #{ app_path } -t 1:1 -w 5 --preload --debug -p #{ port } > #{puma_log}")
-    sleep 5
-    assert_match "PumaWorkerKiller:", puma_log.read
-  ensure
-    puma_log.delete
-    Process.kill('TERM', pid) if pid
-  end
-
-  def test_worker_reaped
-    ram      = 1 #mb
-    cluster = FakeCluster.new
-    reaper   = PumaWorkerKiller::Reaper.new(ram, cluster)
-    worker_count = 10
-    worker_count.times { cluster.add_worker }
-
-    assert_equal worker_count, cluster.workers.count
-    refute cluster.workers.detect {|w| w.is_term? }
+    command  = "bundle exec puma #{ fixture_path.join("default.ru") } -t 1:1 -w 2 --preload --debug -p #{ port }"
+    options  = { wait_for: "booted", timeout: 5, env: { "PUMA_FREQUENCY" => 1 } }
 
-    reaper.reap
-    assert_equal 1, cluster.workers.select {|w| w.is_term? }.count
-
-    reaper.reap
-    assert_equal 2, cluster.workers.select {|w| w.is_term? }.count
-
-    reaper.reap
-    assert_equal 3, cluster.workers.select {|w| w.is_term? }.count
-  ensure
-    cluster.workers.map(&:term)
+    WaitForIt.new(command, options) do |spawn|
+      assert_contains(spawn, "PumaWorkerKiller")
+    end
   end
 
-  def test_kills_memory_leak
-    ram     = rand(75..100) #mb
-    cluster = FakeCluster.new
-    reaper  = PumaWorkerKiller::Reaper.new(ram, cluster)
-    while reaper.get_total_memory < (ram * 0.80)
-      cluster.add_worker
-      sleep 0.01
+  def test_without_preload
+    port     = 0 # http://stackoverflow.com/questions/200484/how-do-you-find-a-free-tcp-server-port-using-ruby
+    command  = "bundle exec puma #{ fixture_path.join("default.ru") } -t 1:1 -w 2 --debug -p #{ port } -C #{ fixture_path.join("config/puma_worker_killer_start.rb") }"
+    options  = { wait_for: "booted", timeout: 10, env: { "PUMA_FREQUENCY" => 1 } }
+
+    WaitForIt.new(command, options) do |spawn|
+      assert_contains(spawn, "PumaWorkerKiller")
     end
+  end
 
-    reaper.reap
-    assert_equal 0, cluster.workers.select {|w| w.is_term? }.count
+  def test_kills_large_app
+    file     = fixture_path.join("big.ru")
+    port     = 0
+    command  = "bundle exec puma #{ file } -t 1:1 -w 2 --preload --debug -p #{ port }"
+    options  = { wait_for: "booted", timeout: 5, env: { "PUMA_FREQUENCY" => 1, 'PUMA_RAM' => 1} }
 
-    until reaper.get_total_memory > ram
-      cluster.add_worker
-      sleep 0.01
+    WaitForIt.new(command, options) do |spawn|
+      assert_contains(spawn, "Out of memory")
     end
-
-    reaper.reap
-    assert_equal 1, cluster.workers.select {|w| w.is_term? }.count
-  ensure
-    cluster.workers.map(&:term)
   end
 
+  def assert_contains(spawn, string)
+    assert spawn.wait(string), "Expected logs to contain '#{string}' but it did not, contents: #{ spawn.log.read }"
+  end
 
   def test_rolling_restart
-    ram     = rand(75..100) #mb
-    cluster = FakeCluster.new
-    cluster.add_worker
 
-    worker = cluster.workers.first
-    reaper = PumaWorkerKiller::RollingRestart.new(cluster)
-    reaper.reap(1)
+    file     = fixture_path.join("rolling_restart.ru")
+    port     = 0
+    command  = "bundle exec puma #{ file } -t 1:1 -w 2 --preload --debug -p #{ port }"
+    puts command.inspect
+    options  = { wait_for: "booted", timeout: 15, env: { } }
 
-    assert_equal 1, cluster.workers.select {|w| w.is_term? }.count
-  ensure
-    cluster.workers.map(&:term)
+    WaitForIt.new(command, options) do |spawn|
+      assert_contains(spawn, "Rolling Restart")
+    end
   end
 end
 

data/test/test_helper.rb

@@ -2,70 +2,8 @@ Bundler.require
 
 require 'puma_worker_killer'
 require 'test/unit'
+require 'wait_for_it'
 
 def fixture_path
   Pathname.new(File.expand_path("../fixtures", __FILE__))
 end
-
-# Mock object stand in for Puma::Cluster
-class FakeCluster
-  def initialize
-    @workers = []
-  end
-
-  def wakeup!
-  end
-
-  class Worker
-    attr_accessor :pid
-
-    def initialize(pid)
-      @pid = pid
-    end
-
-    def memory_leak
-      while true
-
-      end
-    end
-
-    # not public interface, used for testing
-    def is_term?
-      @first_term_sent
-    end
-
-    def term
-      begin
-        if @first_term_sent && (Time.new - @first_term_sent) > 30
-          @signal = "KILL"
-        else
-          @first_term_sent ||= Time.new
-        end
-
-        Process.kill "TERM", @pid
-        Process.wait(@pid)
-      rescue Errno::ESRCH, Errno::ECHILD
-      end
-    end
-  end
-
-  def log(msg)
-    puts msg
-  end
-
-  def do_work
-    while true
-      sleep 1
-    end
-  end
-
-  # not a public interface, added to make testing easier
-  def workers
-    @workers
-  end
-
-  def add_worker
-    pid = fork { do_work }
-    @workers << Worker.new(pid)
-  end
-end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: puma_worker_killer
 version: !ruby/object:Gem::Version
-  version: 0.0.6
+  version: 0.0.7
 platform: ruby
 authors:
 - Richard Schneeman
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2016-04-01 00:00:00.000000000 Z
+date: 2016-10-13 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: puma
@@ -58,6 +58,20 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '1.6'
+- !ruby/object:Gem::Dependency
+  name: wait_for_it
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.1'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.1'
 - !ruby/object:Gem::Dependency
   name: rake
   requirement: !ruby/object:Gem::Requirement
@@ -106,7 +120,11 @@ files:
 - lib/puma_worker_killer/rolling_restart.rb
 - lib/puma_worker_killer/version.rb
 - puma_worker_killer.gemspec
-- test/fixtures/app.ru
+- test/fixtures/big.ru
+- test/fixtures/config/puma_worker_killer_start.rb
+- test/fixtures/default.ru
+- test/fixtures/fixture_helper.rb
+- test/fixtures/rolling_restart.ru
 - test/puma_worker_killer_test.rb
 - test/test_helper.rb
 homepage: https://github.com/schneems/puma_worker_killer
@@ -129,12 +147,17 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.5.1
+rubygems_version: 2.6.4
 signing_key: 
 specification_version: 4
 summary: If you have a memory leak in your web code puma_worker_killer can keep it
   in check.
 test_files:
-- test/fixtures/app.ru
+- test/fixtures/big.ru
+- test/fixtures/config/puma_worker_killer_start.rb
+- test/fixtures/default.ru
+- test/fixtures/fixture_helper.rb
+- test/fixtures/rolling_restart.ru
 - test/puma_worker_killer_test.rb
 - test/test_helper.rb
+has_rdoc: