Re: Something makes curl hang for 5 minutes after connection refused

[basinilya@gmail.com]

This looks fixed now. Thanks.

On 15.09.2019 10:07, Basin Ilya wrote:
> Hi.
> Any update on this?
>
> 01.12.2018 12:51, Corinna Vinschen пишет:
>> On Nov 30 20:42, Brian Inglis wrote:
>>> On 2018-11-30 12:49, Lee wrote:
>>>> On 11/30/18, Basin Ilya  wrote:
>>>>> Hi.
>>>>>
>>>>> Recently I noticed that `curl` does not fail immediately after connection
>>>>> refused, but waits for 5 minutes.
>>>>>
>>>>> This only happens on my machine and only with the `Cygwin` version of
>>>>> `curl`. The mingw version is unaffected.
>>>>> I tried downgrading curl, but it didn't help. I thought there's a rogue
>>>>> setting in `/etc` or in home or maybe some environment variable, but I
>>>>> couldn't find it.
>>>>>
>>>>>     basin@BASIN /cygdrive/c/Users/basin
>>>>>     $ >/dev/null /usr/bin/curl -v http://127.0.0.1:2/
>>>>>     * STATE: INIT => CONNECT handle 0x80048658; line 1404 (connection
>>>>> #-5000)
>>>>>     * Added connection 0. The cache now contains 1 members
>>>>>     *   Trying 127.0.0.1...
>>>>>     * TCP_NODELAY set
>>>>>     * STATE: CONNECT => WAITCONNECT handle 0x80048658; line 1456 (connection
>>>>> #0)
>>>>     <.. snip ..>
>>>>>     curl: (28) Connection timed out after 300145 milliseconds
>>>>>
>>>>>     basin@BASIN /cygdrive/c/Users/basin
>>>>>     $ /usr/bin/curl --version
>>>>>     curl 7.59.0 (i686-pc-cygwin) libcurl/7.59.0 OpenSSL/1.0.2p zlib/1.2.11
>>>>> libidn2/2.0.4 libpsl/0.18.0 (+libidn2/2.0.2) libssh2/1.7.0 nghttp2/1.31.0
>>>>> Can someone try to reproduce it?
>>>> I get the same behavior:
>>>> $ /usr/bin/curl -v http://127.0.0.1:2/
>>>> * STATE: INIT => CONNECT handle 0x600057ad0; line 1404 (connection #-5000)
>>>> * Added connection 0. The cache now contains 1 members
>>>> *   Trying 127.0.0.1...
>>>> * TCP_NODELAY set
>>>> * STATE: CONNECT => WAITCONNECT handle 0x600057ad0; line 1456 (connection #0)
>>>> * Connection timed out after 300324 milliseconds
>>>> * multi_done
>>>> * stopped the pause stream!
>>>> * Closing connection 0
>>>> * The cache now contains 0 members
>>>> curl: (28) Connection timed out after 300324 milliseconds
>>>>
>>>> $ /usr/bin/curl --version
>>>> curl 7.59.0 (x86_64-unknown-cygwin) libcurl/7.59.0 OpenSSL/1.0.2p
>>>> zlib/1.2.11 libidn2/2.0.4 libpsl/0.18.0 (+libidn2/2.0.2) libssh2/1.7.0
>>>> nghttp2/1.31.0
>>>>
>>>>
>>>> I also get an almost immediate 'failed to connect' notice on windows:
>>>> C:\>C:\UTIL\curl\curl.exe -v http://127.0.0.1:2/
>>>> *   Trying 127.0.0.1...
>>>> * TCP_NODELAY set
>>>> * connect to 127.0.0.1 port 2 failed: Connection refused
>>>> * Failed to connect to 127.0.0.1 port 2: Connection refused
>>>> * Closing connection 0
>>>> curl: (7) Failed to connect to 127.0.0.1 port 2: Connection refused
>>>>
>>>> with either version of curl:
>>>> C:\>where curl
>>>> C:\UTIL\curl\curl.exe
>>>> C:\Windows\System32\curl.exe
>>> Ditto all the way!
>>>
>>> Could this be caused by network speed improvements made to Cygwin a while back?
>>> Or were those patches never sent/received/applied from Daniel Havey (MS Windows
>>> Program Manager for Transports and IP)?
>> Somebody may want to bisect this...
>>
>>
>> Corinna
>>