Asset serving from Webrick causes mutex error

[richardnuno]

We're seeing a issue related to asset serving with Webrick that is causing the server to lock up.
These errors were intermittent and hard to reproduce but this test application tries to do that.

When refreshing the root page repeatedly the following error appears in the server log (you might have to hold refresh hotkeys for a few seconds):

ERROR Errno::ECONNRESET: Connection reset by peer @ io_fillbuf - fd:16

or

ERROR Errno::EPROTOTYPE: Protocol wrong type for socket

When shutting down the server the follow error is displayed:

Internal Server Error

Attempt to unlock a mutex which is not locked

Removing the requires from application.js seemed to make the Errno errors disappear.

[richardnuno]

Forgot to mention, ruby version is 2.2.0 and rails version is 4.2.0

[sivsushruth]

@richardnuno I tried to reproduce the bug with ruby 2.1.0 and rails 4.2.0 several times, but in vein.

[TiuTalk]

@richardnuno Can you please provide a sample application that reproduces the issue?

[kamui]

@TiuTalk The sample app is here: https://github.com/bonobos/asset_mutex_bug

I can reproduce the issue on my MacBook Pro running 10.10 with ruby 2.2.0.

[binyamindavid]

Having the same problem.

Completed 200 OK in 102ms (Views: 101.1ms | ActiveRecord: 0.0ms)
[2015-02-23 19:58:50] ERROR Errno::ECONNRESET: Connection reset by peer @ io_fillbuf - fd:16
    /Users/benjaminsigidi/.rvm/rubies/ruby-2.2.0/lib/ruby/2.2.0/webrick/httpserver.rb:80:in `eof?'
    /Users/benjaminsigidi/.rvm/rubies/ruby-2.2.0/lib/ruby/2.2.0/webrick/httpserver.rb:80:in `run'
    /Users/benjaminsigidi/.rvm/rubies/ruby-2.2.0/lib/ruby/2.2.0/webrick/server.rb:294:in `block in start_thread'

[danieldocki]

Same here
Ruby 2.1.2
Rails 4.0.12

11:28:10 web.1  | [2015-03-04 11:28:10] ERROR Errno::ECONNRESET: Connection reset by peer @ io_fillbuf - fd:16
11:28:10 web.1  |   /Users/danieldocki/.rbenv/versions/2.1.2/lib/ruby/2.1.0/webrick/httpserver.rb:80:in `eof?'
11:28:10 web.1  |   /Users/danieldocki/.rbenv/versions/2.1.2/lib/ruby/2.1.0/webrick/httpserver.rb:80:in `run'
11:28:10 web.1  |   /Users/danieldocki/.rbenv/versions/2.1.2/lib/ruby/2.1.0/webrick/server.rb:295:in `block in start_thread'

[arthurnn]

This seems like to be a bug on Webrick and not on Rails TBH.

[ossrapaport-zz]

It's a Webrick bug, switch to Thin and it goes away.

[rails-bot]

This issue has been automatically marked as stale because it has not been commented on for at least
three months.

The resources of the Rails team are limited, and so we are asking for your help.

If you can still reproduce this error on the 4-2-stable, 4-1-stable branches or on master,
please reply with all of the information you have about it in order to keep the issue open.

Thank you for all your contributions.

[arthurnn]

This belongs to Webrick as far as I can tell.

[TonyShuangYuan]

I also encounter this issue in Rails 4.2.6, Some times chrome hold refresh hotkeys for several minutes .After i hit keys in rails console(my rails runs as develop mode) the console display :"ERROR Errno::ECONNRESET: An existing connection was forcibly closed by the remote host..."

[ioquatix]

The EPROTOTYPE issue is a bug with the macOS kernel. More details can be found here: https://bugs.ruby-lang.org/issues/14713