Spark as sql engine on S3

[Ashok Kumar]

Hello gurus,
We are storing data externally on Amazon S3
What is the optimum or best way to use Spark as SQL engine to access data on S3?
Any info/write up will be greatly appreciated.
Regards