HDDS-3930. Fix OMKeyDeletesRequest.

[bharatviswa504]

What changes were proposed in this pull request?

1. The cache of the key should be updated in ValidateAndUpdateCache, as we return response once after adding to cache, and before DoubleBuffer flushes to disk using OmClientResponse#addToDBBatch.
2. Changed return type of DeleteKeysResponse, as anyway when success we are not returning delete list and also when failure, we fail complete batch. So we can just return status.

What is the link to the Apache JIRA

https://issues.apache.org/jira/browse/HDDS-3930

Please replace this section with the link to the Apache JIRA)

How was this patch tested?

Added tests

[xiaoyuyao]

Thanks @bharatviswa504 for the update. LGTM, +1 from me.

[adoroszlai]

Thanks @bharatviswa504 for updating the patch. Looks mostly good, only 2 minor comments.

[adoroszlai]

More arguments (not counting exception) than placeholders:

Suggested change

	LOG.error("Keys delete failed. Volume:{}, Bucket:{}, DeletedKey:{}, " +
	"UnDeletedKeys:{}", volumeName, bucketName, keyName,
	auditMap.get(DELETED_KEYS_LIST), auditMap.get(UNDELETED_KEYS_LIST),
	LOG.error("Keys delete failed. Volume:{}, Bucket:{}, DeletedKeys:{}, " +
	"UnDeletedKeys:{}", volumeName, bucketName,
	auditMap.get(DELETED_KEYS_LIST), auditMap.get(UNDELETED_KEYS_LIST),

[adoroszlai]

Nit: log level mismatch (isDebugEnabled vs. error).

[bharatviswa504]

Thank You @adoroszlai for the review.
I have addressed review comments.

[adoroszlai]

Thanks @bharatviswa504 for updating the patch.

[smengcl]

+1

[bharatviswa504]

Thank You @adoroszlai @xiaoyuyao and @smengcl for the review.