Tracking in upstream Cruise Control issue here [0]

The `/kafkacruisecontrol/user_tasks` endpoint returns task information that includes a `StartMs` field [1], which represents the task’s start time as a UTC timestamp in milliseconds since the Unix epoch. This is very useful for calculating metrics related to task duration.

The task would entail including this `StartMs` field in the response from the `/kafkacruisecontrol/state` endpoint when querying the Executor State [2] as well?

Currently, the Executor State response provides the start timestamp embedded as a string in the `triggeredTaskReason` field [3], alongside the reason and client requester IP. Extracting the timestamp from this field requires parsing the full string, which adds extra complexity in client applications. Including a separate `StartMs` field to these responses would simplify consumption and improve consistency between the two endpoints.

This enhancement would help us simplify our implementation of the Kafka Rebalance Tracking feature [4] which currently relies on extracting task start time from the `triggeredTaskReason`.

[0] https://github.com/linkedin/cruise-control/issues/2271
[1] https://github.com/linkedin/cruise-control/blob/main/cruise-control/src/main/java/com/linkedin/kafka/cruisecontrol/servlet/UserTaskManager.java#L636
[2] https://github.com/linkedin/cruise-control/blob/main/cruise-control/src/main/java/com/linkedin/kafka/cruisecontrol/executor/ExecutorState.java
[3] https://github.com/linkedin/cruise-control/blob/main/cruise-control/src/main/java/com/linkedin/kafka/cruisecontrol/executor/ExecutorState.java#L26

[4] https://github.com/strimzi/strimzi-kafka-operator/pull/11348