From: merch-redmine@...
Date: 2021-07-01T19:33:22+00:00
Subject: [ruby-core:104462] [Ruby master Bug#17561] The timeout option for Addrinfo.getaddrinfo is not reliable on Ruby 2.7.2

Issue #17561 has been updated by jeremyevans0 (Jeremy Evans).


usa (Usaku NAKAMURA) wrote in #note-5:
> About 2.7, to make raise `ArgumentError` (or other exception) is not acceptable because it's **stable* branch.
> I strongly hope to fix this problem at master branch and the fix can be backportable, but at this time, we should simply ignore `timeout:` argument and note it on documentation.

I sent a pull request to disable the timeout option: https://github.com/ruby/ruby/pull/4618

I couldn't see where the timeout option was documented, so I didn't modify the documentation.

I don't think we'll be able to backport a fix to the issue.  The use of getaddrinfo_a was completely removed in commit:5d8bcc4870601ab1ee05346346f241d4a805aac9.

----------------------------------------
Bug #17561: The timeout option for Addrinfo.getaddrinfo is not reliable on Ruby 2.7.2
https://bugs.ruby-lang.org/issues/17561#change-92718

* Author: smcgivern (Sean McGivern)
* Status: Assigned
* Priority: Normal
* Assignee: Glass_saga (Masaki Matsushita)
* ruby -v: 2.7.2
* Backport: 2.5: DONTNEED, 2.6: DONTNEED, 2.7: REQUIRED, 3.0: DONTNEED
----------------------------------------
#15553 introduced a timeout option for Addrinfo.getaddrinfo, which uses getaddrinfo_a internally. It appears this has since been reverted in the development branch via https://github.com/ruby/ruby/commit/5d8bcc4870601ab1ee05346346f241d4a805aac9 (due to #17220 maybe; I didn't quite follow the discussion).

However, Ruby 2.7.2 is still not reliable, even without forking:

```
$ ruby -v
ruby 2.7.2p137 (2020-10-01 revision 5445e04352) [x86_64-linux]
$ ruby -e "require 'resolv'; 10000.times { |i| p [i, Addrinfo.getaddrinfo('2130706433', 80, nil, :STREAM, timeout: 5)] }" | tail
Traceback (most recent call last):
	3: from -e:1:in `<main>'
	2: from -e:1:in `times'
	1: from -e:1:in `block in <main>'
-e:1:in `getaddrinfo': getaddrinfo_a: All requests done (SocketError)
[1473, [#<Addrinfo: 127.0.0.1:80 TCP (2130706433)>]]
[1474, [#<Addrinfo: 127.0.0.1:80 TCP (2130706433)>]]
[1475, [#<Addrinfo: 127.0.0.1:80 TCP (2130706433)>]]
[1476, [#<Addrinfo: 127.0.0.1:80 TCP (2130706433)>]]
[1477, [#<Addrinfo: 127.0.0.1:80 TCP (2130706433)>]]
[1478, [#<Addrinfo: 127.0.0.1:80 TCP (2130706433)>]]
[1479, [#<Addrinfo: 127.0.0.1:80 TCP (2130706433)>]]
[1480, [#<Addrinfo: 127.0.0.1:80 TCP (2130706433)>]]
[1481, [#<Addrinfo: 127.0.0.1:80 TCP (2130706433)>]]
[1482, [#<Addrinfo: 127.0.0.1:80 TCP (2130706433)>]]
```

This is on a VirtualBox VM and fails fairly quickly. On a 'real' Linux system, I need to try a few times or bump the number of iterations, but it also fails consistently with consecutive requests. I'm choosing 2130706433 (the decimal representation of 127.0.0.1) as that's what our test suite uses, and that's what failed when I tried to use the timeout option.

On Ruby 3.0.0-dev this does not fail due to the aforementioned revert, but should this also be removed from Ruby 2.7 until it's ready?



-- 
https://bugs.ruby-lang.org/

Unsubscribe: <mailto:ruby-core-request@ruby-lang.org?subject=unsubscribe>
<http://lists.ruby-lang.org/cgi-bin/mailman/options/ruby-core>