Having used another CQRS/ES framework I did take a look at Wolkenkit, too. It looks like a nice framework, with might lack some advanced features, but does model CQRS/ES using simple but smart APIs.
Background: I did read the documentation but did not yet took a look at it in practise.
One point the documentation did not answer though, but is IMHO important is the question how by its architecture Wolkenkit implements horizontal scaling, meaning adding additional services instances (especially by different number for write and read side). It sounds like it should be possible, but it is (IMHO) not explained how and why.
CQRS/ES has a few potential synchronisation/serialization points, in cases where the order of commands (possible handled by optimistic locking) or even more important the order of events for a single aggregate instance has to be guaranteed (for example read side can not be build correctly if events have the wrong order).
I do not see this answered in the documentation and I would assume that the usage of RabbitMQ alone does not solve this problem.
Has it has been solved explicitly (by a custom infrastructure element?) or are there some (unmentioned) constrains, that (do not) solve this implicitly? Simple link to documentation is fine if I missed something