> Why must we track authentication failures? Shouldn't the calling code handle > this via its logic?
@andrewgaul: From what I recall (and @zack-shoylev will hopefully correct me where I'm wrong), this is not about "hard" 401s, it's about transient authentication failures caused by expiring credentials. Handling that on the caller side would evidently mean wrapping many if not most calls made by the caller with "try and renew creds on failure" logic. Rather than have that all over the place on the caller side, the idea was to handle this within the provider itself. --- Reply to this email directly or view it on GitHub: https://github.com/jclouds/jclouds/pull/104#issuecomment-24474035
