The CRM Middleware Repository 
The CRM Middleware Repository
Purpose
A repository is a collection of tables containing meta-data. A repository object is a subset of meta-data.
The CRM Middleware Repository is used to store meta-level information for repository object types.
The repository data is maintained and stored in the CRM Server.
All meta-data is stored in the the CRM Middleware Repository, which actually consists of several repositories. The repository objects are stored in these repositories.
This document describes the available functions and tools:
The meta-level information in the repository is used to generate runtime-level objects such as function module definitions for interfaces and services.
Runtime-level objects must be initially created for the object types shipped to a customer or re-created after a customer modification affecting the meta-level information.
The BDoc meta-objects are maintained with the BDoc Modeler.
Integration
The components of the CRM Middleware repository can be represented in a layer architecture consisting of the following:
The repository objects of the CRM repository should always be modified using repository interfaces. The repository interfaces are RFC-enabled function modules that call the corresponding repository services, which ensure the consistency and validity of the repository and the middleware server runtime environment.
The Meta-Object Protocol defines the sequential activities a repository service has to perform. There are different types of MOPs depending on the repository object type and the activity type.
This layer consists of the following elements:
The CRM repository provides for a general versioning of repository objects.
It has the following features:
If repository data is relevant for transport into other CRM Servers, functions of the R/3 Change and Transport System are called to create, modify and release transport requests for the repository data. Dependent objects are put into the same transport request.
Only meta-objects are transported between CRM Servers. Generated data is not transported. Instead, it is regenerated locally on the target CRM Server.
Some repository object types need an after-import method. It carries out additional activities during the import of the repository object into the target system. An after-import method is a function module assigned to the corresponding transport object.
After-import methods have the following functionality:
The generation component provides a framework for registering meta-info functions for a specific repository object type. Each time the CRM Middleware Repository data is changed, the runtime objects are regenerated depending on the changed meta-object types. This ensures that the R/3 Runtime Repository data is always consistent with the meta-objects. The generation component also creates the service-specific data and additional BDoc type meta-information, like mapping between BDoc type segments and database tables. This data is needed to generate services.
The modifications of BDoc types and dependent objects must be replicated to any subscribed sites
Replication is modeled by the administrator using the Administration Console. For more information, see the
The Repository Access Layer provides the following functionality:
The repository objects are stored in different repositories. The objects must be modified in order to adapt the BDoc message processing in the CRM Server. Each repository provides basic interfaces to access its objects. The CRM Middleware Repository controls and centralizes the modification of all repository objects. This ensures the consistency and validity of the entire CRM Middleware Repository.