spark-sql migration

spark-sql migration

Since Spark 3.2, Parquet files with nanosecond precision for timestamp type (INT64 (TIMESTAMP(NANOS, true))) are not readable. To restore the behavior before Spark 3.2, you can set spark.sql.legacy.parquet.nanosAsLong to true.

https://github.com/apache/spark/blob/fa66d28f0e34342b72706210d685180c084938d9/docs/sql-migration-guide.md?plain=1#L179