> Might these have been better placed as operations within a single > service? In other words, you didn't have 3 different credit risk > services, you had 1 credit risk service with multiple operations. > One credit service that defined 42 data elements which were exposed > in whole or in part by the service operations. > > This may seem pedantic but I think the distinction is valuable. The > credit risk data/data model wasn't shared amongst services, but > managed by a single service, manipulated by various operations > within that service.
I'd set the boundary higher though, whether they are operations on one service or multiple services (or multiple resources for that matter) would be less important to me than whether I considered them part of the same higher level business service. Thats probably ultra vague but what I mean is that if I was using the sort of approach from Steve Jones' book and in my discussion with the business I find out that they consider Credit Risk as part of a higher level Risk capability/business service then that might become the logical boundary. We might then find a number of Credit Risk service endpoints/REST resources but we would at least have a good idea of how they fit into the bigger picture. This is the side of SOA that makes most sense to me anyway. Actually would it be fair to say that granularity and the whole canonical model versus specific models (data or domain) are the two issues that people on here are most split on? - Colin
