@prakash573: I

I guess spark uses "Snappy" compression for parquet file by default. I'm referring Spark's official document "Learning Spark" , Chapter 9, page # 182, Table 9-3.

Please confirm if this is not correct.

Thank You

Venkat Anampudi