Lighter session changed state as ERROR 30 mins after pod deletion, despite 5s LIGHTER_SESSION_TRACK_RUNNING_INTERVAL

[HarshithaRmAkamai]

Here is our configuration for LIGHTER_SESSION_TRACK_RUNNING_INTERVAL set to 5 seconds:

• name: LIGHTER_SESSION_TRACK_RUNNING_INTERVAL
  value: "5s"
  This setting determines how frequently tasks update session states such as 'ready', 'idle', 'killed', and others.

During testing:
I deleted the driver pod to observe state updates, expecting the session state to change to 'killed' within 5 seconds. However, it actually took nearly 30 minutes for the state to update, and it eventually showed as state as ERROR.

Expectation 1: We anticipated the session state to transition within 5 seconds after deleting the pod.
Expectation 2: We expected the eventual state to be 'KILLED', not 'ERROR' after the pod deletion

Actual : It took nearly 30 minutes for the state transition to 'ERROR'.

We would appreciate assistance in understanding how the lighter configurations influence session states and which actions trigger these state updates.