in S3AFileSystem.initialize(), we check for the bucket existing with verifyBucketExists(), which calls s3.doesBucketExist(). But that doesn't check for auth issues.
s3. doesBucketExistV2() does at least validate credentials, and should be switched to. This will help things fail faster
(this is a dupe of
HADOOP-15409; moving off git PRs so we can get yetus to test everything)