Method call gives an error "HTTPClient::ConnectTimeoutError" with different adapters

[yaairfernando]

Bug report

Current behavior:
I get an error when calling a method, I am not sure why I am getting this error but I think is related to the adapter that as far as I know httpclient is the default adapter. I have tried to set a different adapter like excon but I get the same error. I hope you can help me out.

I have also tried to add some attributes like, open_timeout: 100, read_timeout: 100, raise_errors: false, follow_redirects: true, host: "http://localhost:3000", but I still get the error.

I also tried with different versions of savon 2.11.1, 2.3.3, 2.12.0

Steps to reproduce current behavior:

  WSDL_URL = 'https://www.superfinanciera.gov.co/SuperfinancieraWebServiceTRM/TCRMServicesWebService/TCRMServicesWebService?WSDL'
  @client = Savon.client(wsdl: WSDL_URL, adapter: :net_http)  # I have also tried with excon and httpclient
  response = @client.call(:query_tcrm)

This code gives me this error depending on the adapter.

.rbenv/versions/2.6.3/lib/ruby/gems/2.6.0/gems/socksify-1.7.1/lib/socksify.rb:178:in `initialize': execution expired (HTTPClient::ConnectTimeoutError)

.rbenv/versions/2.6.3/lib/ruby/gems/2.6.0/gems/excon-0.76.0/lib/excon/socket.rb:285:in `select_with_timeout': connect_write timeout reached (Excon::Error::Timeout)

Net::OpenTimeout (execution expired):

Expected behavior:
The method call should be successful.

System information:

• ruby version: '2.6.4'
• savon version: '2.12.0'

[olleolleolle]

Here's a runnable example, with the failure described

require 'bundler/inline'

gemfile do
  source 'https://rubygems.org'
  gem 'savon'
end

WSDL_URL = 'https://www.superfinanciera.gov.co/SuperfinancieraWebServiceTRM/TCRMServicesWebService/TCRMServicesWebService?WSDL'
pp client = Savon.client(wsdl: WSDL_URL) 
pp client.operations
pp response = client.call(:query_tcrm)

pp response

Diagnostic question:
Can you use SmartBear's "SoapUI Open Source" to call this Web Service operation? https://www.soapui.org/downloads/soapui/

I tried with this, too, no change:

  gem 'wasabi', '= 3.5.0'  # Here, I experimented, used an earlier release of Wasabi, a dependency
                                               # didn't work with that, either

[yaairfernando]

I tried running the Web Service with Python and it worked. So I know the web service is working. I also used SoapUI Open Source.

So there is a problem with the Web Service or with the configurations on my side when using the gem?

[marcosdemelo]

@YairFernando67 Have you seen this below? Could you check if it solves your problem and report here?

ruby/ruby#597 (comment)

[yaairfernando]

@marcosdemelo I tried it but it didn't work.

I also try adding env_namespace: 'soap', but didn't work either.

@client = Savon.client(wsdl: WSDL_URL, env_namespace: 'soap')

[marcosdemelo]

@YairFernando67 are you using the certificate that they recommend in the documentation?

Search it on Google: superfinanciera wsdl pdf

20151228manualserviciowebtrm.pdf

[yaairfernando]

@marcosdemelo yes I am using the certificate but I get the following errors.

  file = File.open('lib/cert.cer', "r")
  @client = Savon.client(wsdl: WSDL_URL, ssl_cert_file: file.read, ssl_cert_key_file: file.read, namespace_identifier: 'xw', 
  env_namespace: 'soap')
  response = @client.call(:query_tcrm)
  p response
rescue HTTPI::SSLError => e
  p e.message

httpi-2.4.5/lib/httpi/auth/ssl.rb:128:in `read': File name too long @ rb_sysopen - -----BEG (Errno::ENAMETOOLONG)

And if I pass the certificate like this.

  @client = Savon.client(wsdl: WSDL_URL, ssl_cert_file: 'lib/cert.cer', ssl_cert_key_file: 'lib/cert.cer', namespace_identifier: 'xw', 
  env_namespace: 'soap')
  response = @client.call(:query_tcrm)
  p response
rescue HTTPI::SSLError => e
  p e.message

I get this other error.

httpi-2.4.5/lib/httpi/auth/ssl.rb:144:in `read': Could not parse PKey: no start line (OpenSSL::PKey::PKeyError)

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[clushie]

I'm also affected by this and can't figure out a way to get https working.

My gut-feeling is that the url is passed at a later stage than the httpi request initialization when using wsdl which will HTTPI (or any other library) attempt to connect to a non ssl-host even though https is defined? But I might be completely wrong with that assumption.

PS: I'm using basic auth, no client/auth certificates are needed.

This is where I'd expect to tell HTTPI to use ssl:

savon/lib/savon/request.rb

Line 9 in d871e3b

@http_request = http_request || HTTPI::Request.new

Update: Turns out the underlying problem was the SOAP document itself which contained a hard-coded port that resulted in these timeouts

Checking the https://www.superfinanciera.gov.co/SuperfinancieraWebServiceTRM/TCRMServicesWebService/TCRMServicesWebService?WSDL URL provided in the example above you'll notice the http://10.172.14.199:8183/SuperfinancieraWebServiceTRM/TCRMServicesWebService/TCRMServicesWebService. Running the minimal-example on ruby 2.5 you'll get a more helpful timeout message too:

Traceback (most recent call last):
16: from ruby.rb:11:in `<main>'
15: from /usr/local/bundle/gems/savon-2.12.1/lib/savon/client.rb:36:in `call'
14: from /usr/local/bundle/gems/savon-2.12.1/lib/savon/operation.rb:54:in `call'
13: from /usr/local/bundle/gems/savon-2.12.1/lib/savon/operation.rb:94:in `call_with_logging'
12: from /usr/local/bundle/gems/savon-2.12.1/lib/savon/request_logger.rb:12:in `log'
11: from /usr/local/bundle/gems/savon-2.12.1/lib/savon/operation.rb:94:in `block in call_with_logging'
10: from /usr/local/bundle/gems/httpi-2.4.5/lib/httpi.rb:133:in `post'
 9: from /usr/local/bundle/gems/httpi-2.4.5/lib/httpi.rb:161:in `request'
 8: from /usr/local/bundle/gems/httpi-2.4.5/lib/httpi/adapter/net_http.rb:38:in `request'
 7: from /usr/local/bundle/gems/httpi-2.4.5/lib/httpi/adapter/net_http.rb:89:in `do_request'
 6: from /usr/local/lib/ruby/2.5.0/net/http.rb:909:in `start'
 5: from /usr/local/lib/ruby/2.5.0/net/http.rb:920:in `do_start'
 4: from /usr/local/lib/ruby/2.5.0/net/http.rb:935:in `connect'
 3: from /usr/local/lib/ruby/2.5.0/timeout.rb:103:in `timeout'
 2: from /usr/local/lib/ruby/2.5.0/timeout.rb:93:in `block in timeout'
 1: from /usr/local/lib/ruby/2.5.0/net/http.rb:936:in `block in connect'
/usr/local/lib/ruby/2.5.0/net/http.rb:939:in `rescue in block in connect': Failed to open TCP connection to 10.172.14.199:8183 (Network is unreachable - connect(2) for "10.172.14.199" port 8183) (Errno::ENETUNREACH

For us the fix was telling our third-party API to change in their Microsoft SOAP settings to the correct soap:address location, but seeing several people are affected by this maybe support for overwriting the soap address host/port/url would be helpful (in case that is not possible yet).

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.