Setting a timeout for Ruby SOAP clients

Last night I was wrestling with a Ruby SOAP client of a long-running web-service. After 60 seconds, the script would exit with the following stack:

      
      /usr/local/lib/ruby/1.8/timeout.rb:54:in `rbuf_fill': execution expired (Timeout::Error)
        from /usr/local/lib/ruby/1.8/timeout.rb:56:in `timeout'
        from /usr/local/lib/ruby/1.8/timeout.rb:76:in `timeout'
        from /usr/local/lib/ruby/1.8/net/protocol.rb:132:in `rbuf_fill'
        from /usr/local/lib/ruby/1.8/net/protocol.rb:116:in `readuntil'
        from /usr/local/lib/ruby/1.8/net/protocol.rb:126:in `readline'
        from /usr/local/lib/ruby/1.8/net/http.rb:2017:in `read_status_line'
        from /usr/local/lib/ruby/1.8/net/http.rb:2006:in `read_new'
        from /usr/local/lib/ruby/1.8/net/http.rb:1047:in `request'
         ... 9 levels...
        from /usr/local/lib/ruby/1.8/soap/rpc/proxy.rb:141:in `call'
        from /usr/local/lib/ruby/1.8/soap/rpc/driver.rb:178:in `call'
        ... snip
      
  

Googling "ruby SOAP timeout" etc yields very little in the way of help. So either everyone is writing those speedy little stock ticker webservice clients I read so much about, or the brain dead should be capable of easily configuring a client timeout. Alas.

After getting a bit of hint from Google Groups and clicking around in Ruby-Doc.org, I found the SOAP::HTTPConfigLoader module. It's set_options method consists of registering a long list of blocks for handling various configuration parameters and it concludes with this:

      
        options.add_hook("connect_timeout") do |key, value|
          client.connect_timeout = value
        end
        options.add_hook("send_timeout") do |key, value|
          client.send_timeout = value
        end
        options.add_hook("receive_timeout") do |key, value|
          client.receive_timeout = value
        end
      
  

I'm still not clear on the relationship between a SOAP::RPC::Driver instance and SOAP::HTTPConfigLoader, and it sure would be nice if a complete set of available configuration parameters were published somewhere outside of the std library source code (or maybe it's out there and I just haven't found it, or maybe I'm failing to appreciate the beauty of all this, um, self-documenting code...). In any event, SOAP::RPC::Driver does have a SOAP::Property member named "options", which can be treated as an associative array. So I give this a try to see if I can set an indefinite timeout:

      
require 'soap/rpc/driver'

client = SOAP::RPC::Driver.new("http://domain/?wsdl", "ServiceName")

client.add_method('method_name', 'arg-name')
client.options["receive_timeout"] = -1

result = client.method_name("some-value")
puts result
    
  

No such luck. Browsing /usr/local/lib/ruby/1.8/timeout.rb, I don't see an obvious way to set an indefinite timeout, and I see from the Timeout::timeout method argument names that i should probably specify my timeout in seconds. So, I change receive_timeout again:

      
client.options["receive_timeout"] = 360
    
  

Success.

So, maybe someone more expert than I with Ruby would have seen the solution faster than I did, or know of a better documentation source, but I did think it was interesting that this problem would have been completely unsolvable without either a debugger or direct access to the std library source code. Is this how Ruby development is supposed to work, i.e. that one needs the source in order to understand a library? I get that this might be consistent with Ruby's affinity for "openness" in all its various manifestations, and sure, I guess I can imagine a techie utopia wherein documentation, configuration and code are indistinguishable. But geez... so much time wasted just to set a timeout....