I wanted to get some insights on the current best practices when it comes to different bounded contexts interacting with each other. As an example, I want to use the following situation:
The bounded context
Employees deals with the complete lifecycle of an employee within an organization. This BC contains the aggregates
Qualification. Qualifications might be assigned to employees. Trivial commands for each aggregate might be:
The second BC is called
DutyPlanning. This BC contains the aggregates
Duty. A worker is someone who is allowed to be assigned to duties. All employees of an organization are also workers. In order to be allowed to be assigned to a duty, the worker must fulfill the minimum qualification of the duty.
DutyQualifications are identical to the qualifications in the BC Employees.
My questions mainly focus on how to keep the BCs in sync and how to implement it in Axon Framework. I have looked at this repository which shows an example of how to achieve this. Here sagas are used to send commands and receive events between BCs.
(1) Do we consider the public commands and events of a BC as its API? I feel that depending on the event schema of another BC might be problematic. How could an ACL mitigate this and how would this ACL be implemented?
In the case of the Qualifications in my example, the two BCs need to be kept in perfect sync regarding changes to the Qualification aggregate.
(2) Is this pattern common/desirable and how could this be implemented? Using sagas seems overkill in this scenario because no state is required. It’s just a matter of sending the appropriate commands to the DutyPlanning BC after an event occurred. External events handlers could be used for this. In which BC would I place these handlers?
Feel free to raise your own questions and expand on my simple example. I highly appreciate any input on this topic