bugfix: list runtimeclass and csidriver from cache error when cloud-edge network disconnected

[rambohe-ch]

Ⅰ. Describe what this PR does

• background:

1. At present, we use component/resource/namespace/name as file path to store response objects from kube-apiserver on local disk.
2. If response objects from kube-apiserver are empty(like kubelet list runtimeclass, csidriver), cache info will not be created on local disk.
3. when cloud-edge network becomes disconnected, kubelet will list runtimeclass, csidriver from local disk cache, and yurthub return 404 to kubele. so kubelet will re-list runtimeclass and csidriver frequently.

• solution:

1. if response objects from kube-apiserver are empty(like kubelet list runtimeclass, csidriver), yurthub can create key info component/resource in local disk with nil data.
2. so when kubelet list runtimeclass, csidriver from local disk cache, yurthub will return empty list to kubelet.
3. and when pods cache data on local disk is deleted by accident, yurthub will return 404 error(not empty list) to kubelet in order to prevent killing pods on node.

Ⅱ. Does this pull request fix one issue?

Ⅲ. List the added test cases (unit test/integration test) if any, please explain if no tests are needed.

=== RUN TestCacheResponseForList/cache_response_for_list_nodes_with_fieldselector
=== RUN TestCacheResponseForList/cache_response_for_list_runtimeclasses_with_no_objects
=== RUN TestCreate/create_dir_key_with_no_data
=== RUN TestCreate/create_dir_key_that_already_exists

Ⅳ. Describe how to verify it

make all
and make sure when cloud-edge networ disconnected, kubelet list runtimeclass and no error has happened.

Ⅴ. Special notes for reviews