fix(engine): route external wakeup callbacks through FlowOperation for correct node targeting

[derek-miller]

Pull Request type

• Bugfix

NOTE: Please remember to run ./gradlew spotlessApply to fix any format violations.

Changes in this PR

FlowEngineController was injecting FlowExecutor and calling it directly for all three endpoints (startFlow, single wakeUp, bulk wakeUp). This bypassed RestBasedFlowOperation — the routing layer that looks up group ownership and forwards requests to the correct node.

The consequence: when an external system (e.g. Relay) fires an HTTP callback to the Kubernetes service DNS name, K8s round-robins the request to any pod. If that pod doesn't own the target group, FlowExecutor.wakeUp() silently returns false and the step falls back to its polling reconciliation cycle.

The internal wakeup path (InstanceActionJobEventProcessor / UpdateJobEventProcessor) already correctly goes through FlowOperation → RestBasedFlowOperation. This fix makes the external HTTP callback path consistent with that.

Fix: replace FlowExecutor injection in FlowEngineController with FlowOperation. RestBasedFlowOperation handles the routing:

1. Looks up the group owner from the DB
2. If local → executes directly via FlowExecutor
3. If remote → forwards to the correct pod's address (no infinite loop: on the second hop the group is local)

The bulk wakeUp endpoint is also simplified from a stream of per-ref FlowExecutor calls to a single flowOperation.wakeUp(groupId, refs, code), doing one DB lookup for the whole batch.

New tests in RestBasedFlowOperationTest verify that when a group is owned by a remote pod, wakeUp (single and bulk) forwards via RestTemplate and never touches the local FlowExecutor.