pitchfork 0.1.2 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 7e837b05925916feb0b10fb28b2ddfbc8bb476aa77091217cbb012fd0c98f6c0
-  data.tar.gz: 5b89fafb322999c204408a642e39547a652b61968bd2eae4b1c87b77ae51baec
+  metadata.gz: f9f15127612a30958ee43b9667990a0728b5fdfc46e13285bdfa1706085eb67d
+  data.tar.gz: '08ba863ccf209644093b83ee905305615cfef046c4cd5e11e74f16540c722ad7'
 SHA512:
-  metadata.gz: df56d4c6bb70b5735e4859c1186b3e754210b84dfd457f804d6f93f2ac804205f351bd2ef557a09f05fa39896bb29180c6974ef9c74a062c8ddbb32506bc5c36
-  data.tar.gz: 93791c5474ec82f7470270f371df480ff7178ab903ee59799ddd975d0154841234895bce720c911b0fafba78016f36cc7fbac7210b7025417bff5fde898da0b8
+  metadata.gz: 6c73e64f686acde16c5bb50df83a3a8596d9a3f578318cdc55bc3d8dbee1a0d7dd8534e627b5baa4994f3f30d4917ffe7d4e5d405850d068f7523cf52c21113b
+  data.tar.gz: 80c872414fe8289c7c98cb9949309f76aeec63b3f3de85f09349839f873721cdb41c12bfa4620c7863d62526533c7e4ee900b7f5461917d348fa79d4c5977e4a

data/.github/workflows/ci.yml

@@ -11,7 +11,7 @@ jobs:
       matrix:
         os: ["ubuntu-latest"]
         redis: ["6.2"]
-        ruby: ["3.1", "3.0", "2.7", "2.6"]
+        ruby: ["3.2", "3.1", "3.0", "2.7", "2.6"]
     runs-on: ubuntu-latest
     steps:
       - name: Check out code

data/CHANGELOG.md

@@ -1,5 +1,12 @@
 # Unreleased
 
+- Remove default middlewares.
+- Refork indefinitely when `refork_after` is set, unless the last element is `false`.
+- Remove `mold_selector`. The promotion logic has been moved inside workers (#38).
+- Add the `after_promotion` callback.
+- Removed the `before_fork` callback.
+- Fork workers and molds with a clean stack to allow more generations. (#30)
+
 # 0.1.2
 
 - Improve Ruby 3.2 and Rack 3 compatibility.

data/Dockerfile

@@ -1,4 +1,4 @@
-FROM ruby:3.1
+FROM ruby:3.2
 RUN apt-get update -y && apt-get install -y ragel socat netcat smem apache2-utils
 WORKDIR /app
 CMD [ "bash" ]

data/Gemfile.lock

@@ -1,26 +1,27 @@
 PATH
   remote: .
   specs:
-    pitchfork (0.1.2)
+    pitchfork (0.2.0)
       rack (>= 2.0)
       raindrops (~> 0.7)
 
 GEM
   remote: https://rubygems.org/
   specs:
-    minitest (5.16.3)
+    minitest (5.15.0)
     nio4r (2.5.8)
-    puma (5.6.5)
+    puma (6.1.1)
       nio4r (~> 2.0)
-    rack (3.0.4.1)
-    raindrops (0.20.0)
+    rack (3.0.7)
+    raindrops (0.20.1)
     rake (13.0.6)
-    rake-compiler (1.2.0)
+    rake-compiler (1.2.1)
       rake
 
 PLATFORMS
   aarch64-linux
   arm64-darwin-21
+  arm64-darwin-22
   x86_64-linux
 
 DEPENDENCIES

data/Rakefile

@@ -11,6 +11,14 @@ Rake::TestTask.new("test:unit") do |t|
   t.warning = true
 end
 
+Rake::TestTask.new("test:integration") do |t|
+  t.libs << "test"
+  t.libs << "lib"
+  t.test_files = FileList["test/integration/**/test_*.rb"]
+  t.options = '-v' if ENV['CI'] || ENV['VERBOSE']
+  t.warning = true
+end
+
 namespace :test do
   # It's not so much that these tests are slow, but they tend to fork
   # and/or register signal handlers, so they if something goes wrong
@@ -32,7 +40,7 @@ namespace :test do
   # It's quite hard to work with and it would be good to convert all this
   # to Ruby integration tests, but while pitchfork is a moving target it's
   # preferable to edit the test suite as little as possible.
-  task integration: :compile do
+  task legacy_integration: :compile do
     File.write("test/integration/random_blob", File.read("/dev/random", 1_000_000))
     lib = File.expand_path("lib", __dir__)
     path = "#{File.expand_path("exe", __dir__)}:#{ENV["PATH"]}"
@@ -67,6 +75,6 @@ task :ragel do
   end
 end
 
-task test: %i(test:unit test:slow test:integration)
+task test: %i(test:unit test:slow test:integration test:legacy_integration)
 
 task default: %i(ragel compile test)

data/docs/CONFIGURATION.md

@@ -238,20 +238,27 @@ application that are used in hooks.
 `pitchfork` also don't attempt to rescue hook errors. Raising from a worker hook will crash the worker,
 and raising from a master hook will bring the whole cluster down.
 
-### `before_fork`
+### `after_promotion`
 
 ```ruby
-before_fork do |server, worker|
+after_promotion do |server, mold|
   Database.disconnect!
 end
 ```
 
-Called in the context of the parent or mold.
+Called in the context of the mold when initially spawned or promotted.
+
+It's usage is similar to a `before_fork` callback found on other servers
+but it is called once on promotion rather than before forking each worker.
+
 For most protocols connections can be closed after fork, but some
 stateful protocols require to close connections before fork.
 
 That is the case for instance of many SQL databases protocols.
 
+This is also the callback in which memory optimizations, such as
+heap compaction should be done.
+
 ### `after_fork`
 
 ```ruby
@@ -263,6 +270,19 @@ end
 Called in the worker after forking. Generally used to close inherited connections
 or to restart backgrounds threads for libraries that don't do it automatically.
 
+### `after_promotion`
+
+```ruby
+after_promotion do |server, mold|
+  NetworkClient.disconnect!
+  4.times { GC.start } # promote surviving objects to oldgen
+  GC.compact
+end
+```
+
+Called in the worker after it was promoted into a mold. Generally used to shutdown
+open connections and file descriptors, as well as to perform memory optimiations
+such as compacting the heap, trimming memory etc.
 
 ### `after_worker_ready`
 
@@ -302,9 +322,15 @@ once at least one worker processed `50` requests.
 
 Each element is a limit for the next generation. On the example above a new generation
 is triggered when a worker has processed 50 requests, then the second generation when
-a worker from the new generation processed an additional 100 requests and finally after another
+a worker from the new generation processed an additional 100 requests and finally after *every*
 1000 requests.
 
+If you don't want unlimited reforking, you can set `false` as the last element of the array:
+
+```ruby
+refork_after [50, 100, 1000, false]
+```
+
 Generally speaking Copy-on-Write efficiency tend to degrade fast during the early requests,
 and then less and less frequently.
 
@@ -314,36 +340,8 @@ By default automatic reforking isn't enabled.
 
 Make sure to read the [fork safety guide](FORK_SAFETY.md) before enabling reforking.
 
-### `mold_selector`
-
-Sets the mold selector implementation.
-
-```ruby
-mold_selector do |server|
-  candidate = server.children.workers.sample # return an random worker
-  server.logger.info("worker=#{worker.nr} pid=#{worker.pid} selected as new mold")
-  candidate
-end
-```
-
-The has access to `server.children` a `Pitchfork::Children` instance.
-This object can be used to introspect the state of the cluster and select the most
-appropriate worker to be used as the new mold from which workers will be reforked.
-
-The default implementation selects the worker with the least
-amount of shared memory. This heuristic aim to select the most
-warmed up worker.
-
-This should be considered a very advanced API and it is discouraged
-to use it unless you are confident you have a clear understanding
-of pitchfork's architecture.
-
 ## Rack Features
 
-### `default_middleware`
-
-Sets whether to add Pitchfork's default middlewares. Defaults to `true`.
-
 ### `early_hints`
 
 Sets whether to enable the proposed early hints Rack API. Defaults to `false`.

data/examples/pitchfork.conf.rb

@@ -23,9 +23,9 @@ check_client_connection false
 # local variable to guard against running a hook multiple times
 run_once = true
 
-before_fork do |server, worker|
+after_promotion do |server, worker|
   # the following is highly recommended for Rails
-  # as there's no need for the master process to hold a connection
+  # as there's no need for the mold process to hold a connection
   defined?(ActiveRecord::Base) and
     ActiveRecord::Base.connection.disconnect!
 

data/exe/pitchfork

@@ -6,7 +6,6 @@ require 'optparse'
 ENV["RACK_ENV"] ||= "development"
 rackup_opts = Pitchfork::Configurator::RACKUP
 options = rackup_opts[:options]
-set_no_default_middleware = true
 
 op = OptionParser.new("", 24, '  ') do |opts|
   cmd = File.basename($0)
@@ -59,11 +58,6 @@ op = OptionParser.new("", 24, '  ') do |opts|
     ENV["RACK_ENV"] = e
   end
 
-  opts.on("-N", "--no-default-middleware",
-          "do not load middleware implied by RACK_ENV") do |e|
-    rackup_opts[:no_default_middleware] = true if set_no_default_middleware
-  end
-
   opts.on("-s", "--server SERVER",
           "this flag only exists for compatibility") do |s|
     warn "-s/--server only exists for compatibility with rackup"
@@ -101,7 +95,6 @@ op = OptionParser.new("", 24, '  ') do |opts|
   opts.parse! ARGV
 end
 
-set_no_default_middleware = false
 app = Pitchfork.builder(ARGV[0] || 'config.ru', op)
 op = nil
 
@@ -113,4 +106,4 @@ if $DEBUG
   })
 end
 
-Pitchfork::HttpServer.new(app, options).start.join
+exit(Pitchfork::HttpServer.new(app, options).start(false).join)

data/ext/pitchfork_http/epollexclusive.h

@@ -64,9 +64,8 @@ static VALUE prep_readers(VALUE cls, VALUE readers)
 
 #if USE_EPOLL
 struct ep_wait {
-	struct epoll_event *events;
+	struct epoll_event event;
 	rb_io_t *fptr;
-	int maxevents;
 	int timeout_msec;
 };
 
@@ -74,8 +73,14 @@ static void *do_wait(void *ptr) /* runs w/o GVL */
 {
 	struct ep_wait *epw = ptr;
 
-	return (void *)(long)epoll_wait(epw->fptr->fd, epw->events,
-				epw->maxevents, epw->timeout_msec);
+	/*
+	 * Linux delivers epoll events in the order received, and using
+	 * maxevents=1 ensures we pluck one item off ep->rdllist
+	 * at-a-time (c.f. fs/eventpoll.c in linux.git, it's quite
+	 * easy-to-understand for anybody familiar with Ruby C).
+	 */
+	return (void *)(long)epoll_wait(epw->fptr->fd, &epw->event, 1,
+					epw->timeout_msec);
 }
 
 /* :nodoc: */
@@ -84,14 +89,10 @@ static VALUE
 get_readers(VALUE epio, VALUE ready, VALUE readers, VALUE timeout_msec)
 {
 	struct ep_wait epw;
-	long i, n;
-	VALUE buf;
+	long n;
 
 	Check_Type(ready, T_ARRAY);
 	Check_Type(readers, T_ARRAY);
-	epw.maxevents = RARRAY_LENINT(readers);
-	buf = rb_str_buf_new(sizeof(struct epoll_event) * epw.maxevents);
-	epw.events = (struct epoll_event *)RSTRING_PTR(buf);
 	epio = rb_io_get_io(epio);
 	GetOpenFile(epio, epw.fptr);
 
@@ -99,17 +100,12 @@ get_readers(VALUE epio, VALUE ready, VALUE readers, VALUE timeout_msec)
 	n = (long)rb_thread_call_without_gvl(do_wait, &epw, RUBY_UBF_IO, NULL);
 	if (n < 0) {
 		if (errno != EINTR) rb_sys_fail("epoll_wait");
-		n = 0;
-	}
-	/* Linux delivers events in order received */
-	for (i = 0; i < n; i++) {
-		struct epoll_event *ev = &epw.events[i];
-		VALUE obj = rb_ary_entry(readers, ev->data.u64);
+	} else if (n > 0) { /* maxevents is hardcoded to 1 */
+		VALUE obj = rb_ary_entry(readers, epw.event.data.u64);
 
 		if (RTEST(obj))
 			rb_ary_push(ready, obj);
-	}
-	rb_str_resize(buf, 0);
+	} /* n == 0 : timeout */
 	return Qfalse;
 }
 #endif /* USE_EPOLL */