Hi Mikhael,
I’m currently working on an OpenShift project and ran into this myself as well.
When you scale a Mendix app to more than one replica in Kubernetes, the operator designates one pod as the master (leader) and the others as workers. The master handles tasks that should only run once, such as database schema updates, scheduled events, and certain background processes.
The worker nodes are mainly there for scalability. They handle user requests, APIs, and UI sessions, but do not execute scheduled events or perform schema changes.
In short, the master coordinates and runs exclusive tasks, while workers provide horizontal scalability, and all pods can still serve user traffic.
Tim