Matt Snider

Here's a link to the Routing Service Blog, another Channel 9 video that discusses the Routing Service in general, and the current set of Routing Service Samples and Docs over on MSDN.

It's not out of the box, you're right.  The main reason why is that the Routing Service is designed to be stateless, and most intelligent load balancing methods rely on some knowledge of "how many messages" or "how much traffic" has been sent a certain direction.  While it's possible to build stateless load balancers, it's not immediately apparent how to do this within the Routing Service.

Over at the blog I'm going to be addressing the load balancing question fairly soon.  In the meantime, take a look at the "Advanced Filters" sample over on MSDN.  One of the advanced filters we build is a Round Robin filter.  Note: I don't recommend using this sample, as it has all sorts of issues with locking.  In the blog I'll explain why this isn't the best design (it's included in the sample as an example of a custom filter) and give you a better solution (hint: the Routing Service utilizes a MessageFilterTable<IEnumerable<ServiceEndpoint>>).

Beta builds of the Routing Service documents are up now publicly.  You can find the documentation at http://msdn.microsoft.com/en-us/library/ee517421(VS.100).aspx">http://msdn.microsoft.com/en-us/library/ee517421(VS.100).aspx and the samples are here http://msdn.microsoft.com/en-us/library/dd699773(VS.100).aspx">http://msdn.microsoft.com/en-us/library/dd699773(VS.100).aspx.  Comment heavily if you feel like they don't explain things well.  I also have started publishing a blog at http://blogs.msdn.com/routingrules/default.aspx.