What are two testing scenarios of the chaos engineering principle? (Choose two.)

Maxing out CPU cores on an Elasticsearch cluster simulates resource contention or failure on a critical data component, testing how the system handles stress and degraded performance, which is a key chaos engineering principle.

Removing all users from a version control system is a security or administrative issue, not a test of system resilience under fault conditions typical of chaos engineering.

Executing routines in driver code to emulate I/O errors describes a *method* for injecting a fault, but is less of a high-level testing *scenario* compared to the broader system impacts of options A and E.

Blocking developers' building access relates to developer workflow or security access and does not test the resilience of the application or infrastructure itself under failure conditions.

Unplugging a core switch device simulates a significant network infrastructure failure, directly testing the system's ability to recover, failover, and maintain availability under severe network disruption, a common chaos engineering experiment.