[k8s] Fix `--purge` not cleaning up cluster in stale k8s context #18915

Job	Run time
python-test (3.8, tests/unit_tests)	3m 32s
python-test (3.8, tests/test_api.py)	42s
python-test (3.8, tests/test_cli.py)	53s
python-test (3.8, tests/test_config.py)	38s
python-test (3.8, tests/test_global_user_state.py)	27s
python-test (3.8, tests/test_jobs.py)	31s
python-test (3.8, tests/test_list_accelerators.py)	44s
python-test (3.8, tests/test_optimizer_dryruns.py)	1m 46s
python-test (3.8, tests/test_optimizer_random_dag.py)	1m 8s
python-test (3.8, tests/test_storage.py)	30s
python-test (3.8, tests/test_wheels.py)	32s
python-test (3.8, tests/test_jobs_and_serve.py)	30s
python-test (3.8, tests/test_yaml_parser.py)	29s
	12m 22s

Provide feedback