Better Orchestration Through Folder Structure
Organizing resources into an appropriate structure is a major struggle for most implementors. Unorganized resources makes it difficult to scale. In addition, confusing folder usage as taxonomy instead of data usage places an unnecessary burden on a consumer of a resource. Maximum orchestration can only be achieved through proper semantics.
A folder should contains all the resources that belong to the current data structure. The folder includes all files, resources and/or business data necessary to process a request.
Most of the time, there is only one data resource. Sometimes, there may be more than one data resource for the same composite resource (e.g., one for http and another on https).
A project folder should contain resources attached to a project. The composite resource may be divided into corresponding systems and inserted into a specific folder called by an orchestrator.
The specific folder name under the system folder contains all resources belonging to the current implementation. The specific folder includes all files and resources. For example, the folder will contain the XSLT files, the XSD files, WADL, and/or WSDL files used to call a resource.
An Infrastructure folder should contain all resources that are not specific to a particular project or implementation. The Infrastructure folder includes the XSD files, Query files, shared function files, and the WSDL file used to implement a generic service.