Deterministic Behavior of Container Kill Serial Experiment #5445
Description
Describe in detail the feature/behavior/change you'd like to see.
There should be a configuration available to control or enforce a fixed pod termination order.
During our testing, we noticed that the pods are being terminated in a non deterministic order, meaning the framework does not follow a fixed sequence for killing pods.
Because of this, in some runs the session pod (critical for our signalling KPI recovery) is killed last, which shortens the actual recovery window before KPI verification. This variation in pod kill order is leading to inconsistent KPI results, even though the configured sleep time remains the same.