The Shared Kernel pattern describes a relationship between two bounded contexts and is used on a context map in CML. The Language Semantics page describes all relationship types that are supported in CML.

Syntax

Note that currently two different syntax variants exist. The following code snippets illustrate both variants:

CargoBookingContext [SK]<->[SK] VoyagePlanningContext {
  implementationTechnology = "Java Library"
}

Note that with this syntax (with the arrows <->) it does not matter which bounded context is on which side, since this is a symmetric relationship. If you switch the bounded contexts, it has the same meaning semantically.

CargoBookingContext Shared-Kernel VoyagePlanningContext {
  implementationTechnology = "Java Library"
}

Default Symmetric Relationship

Note that the Shared Kernel is the default symmetric relationship. Omitting the concrete relationship type within the brackets as follows, declares a Shared Kernel relationship too. However, with this syntax a reader has to know this behavior and cannot explicitly see that it is not a Partnership relationship:

CargoBookingContext <-> VoyagePlanningContext

Implementation Technology

With the implementationTechnology keyword you can specify how the relationship is implemented.

Relationship Name

With a colon it is possible (optionally) to add a relationship name to the specification, as illustrated within this example:

CargoBookingContext Shared-Kernel VoyagePlanningContext : BookingVoyageRelationship {
  implementationTechnology = "Java Library"
}