机架中间件死锁

Question

I include this simple Rack Middleware in a Rails application:

class Hello

  def initialize(app)
    @app = app
  end

  def call(env)
    [200, {"Content-Type" => "text/html"}, "Hello"]
  end

end

Plug it in inside environment.rb:

...
Dir.glob("#{RAILS_ROOT}/lib/rack_middleware/*.rb").each do |file|
  require file
end
Rails::Initializer.run do |config|
  config.middleware.use Hello
...

I'm using Rails 2.3.5, Webrick 1.3.1, ruby 1.8.7

When the application is started in production mode, everything works as expected - every request is intercepted by the Hello middleware, and "Hello" is returned. However, when run in development mode, the very first request works returning "Hello", but the next request hangs.

Interrupting webrick while it is in the hung state yields this:

^C[2010-03-24 14:31:39] INFO  going to shutdown ...
deadlock 0xb6efbbc0: sleep:-  - /usr/lib/ruby/gems/1.8/gems/actionpack-2.3.5/lib/action_controller/reloader.rb:31
deadlock 0xb7d1b1b0: sleep:J(0xb6efbbc0) (main) - /usr/lib/ruby/1.8/webrick/server.rb:113
Exiting
/usr/lib/ruby/1.8/webrick/server.rb:113:in `join': Thread(0xb7d1b1b0): deadlock (fatal)
 from /usr/lib/ruby/1.8/webrick/server.rb:113:in `start'
 from /usr/lib/ruby/1.8/webrick/server.rb:113:in `each'
 from /usr/lib/ruby/1.8/webrick/server.rb:113:in `start'
 from /usr/lib/ruby/1.8/webrick/server.rb:23:in `start'
 from /usr/lib/ruby/1.8/webrick/server.rb:82:in `start'
 from /usr/lib/ruby/gems/1.8/gems/rack-1.0.1/lib/rack/handler/webrick.rb:14:in `run'
 from /usr/lib/ruby/gems/1.8/gems/rails-2.3.5/lib/commands/server.rb:111
 from /usr/local/lib/site_ruby/1.8/rubygems/custom_require.rb:31:in `gem_original_require'
 from /usr/local/lib/site_ruby/1.8/rubygems/custom_require.rb:31:in `require'
 from script/server:3

Something to do with the class reloader in development mode. There is also mention of deadlock in the exception.

Any ideas what might be causing this? Any recommendations as to the best approach to debug this?

UPDATE

$ script/console 
Loading development environment (Rails 2.3.5)
>> app.get '/'
=> 200
>> app.get '/'
ThreadError: stopping only thread
 note: use sleep to stop forever
 from /usr/lib/ruby/gems/1.8/gems/actionpack-2.3.5/lib/action_controller/reloader.rb:31:in `lock'
 from /usr/lib/ruby/gems/1.8/gems/actionpack-2.3.5/lib/action_controller/reloader.rb:31:in `run'
 from /usr/lib/ruby/gems/1.8/gems/actionpack-2.3.5/lib/action_controller/dispatcher.rb:108:in `call'
 from /usr/lib/ruby/gems/1.8/gems/rack-1.0.1/lib/rack/lint.rb:47:in `_call'
 from /usr/lib/ruby/gems/1.8/gems/rack-1.0.1/lib/rack/lint.rb:35:in `call'
 from /usr/lib/ruby/gems/1.8/gems/actionpack-2.3.5/lib/action_controller/integration.rb:316:in `process'
 from /usr/lib/ruby/gems/1.8/gems/actionpack-2.3.5/lib/action_controller/integration.rb:197:in `get'
 from (irb):2

Looks like it might be related to this issue:

https://rails.lighthouseapp.com/projects/8994/tickets/3153-actioncontrollerintegrationsession-broken-in-234

Answer 1

I've come up with a hack that will get me by the time being. It's on the ticket mentioned above.

https://rails.lighthouseapp.com/projects/8994/tickets/3153-actioncontrollerintegrationsession-broken-in-234