I know this is a multi-component issue, but can you do the initial triage for this bug. The problem is that when people use the REST API to delete and recreate a bucket they can get error message like this.
Http Error: 500 Reason: Internal Server Error Details: No reason given
Http Error: 503 Reason: Service Unavailable Details: No reason given
I know one issue here has to do with how fast we create vbucket files on the engine side, but in my opinion I think we can also do better on the ns_server side by providing some better error messages. The SDK side may also benefit too from having some good retry code. Can you add your thoughts to the comments section and assign this issue back to be when your done.