Datafactory : Oracle datatime mapping to parquet

asked 2021-11-02T04:16:11.887+00:00
sakuraime 2,271 Reputation points

I have a oracle table field with Date , and while I copy the table to parquet , and import the schema , it help me to convert to int_96 .... why is that? parquet should have date type for data type .

my oracle date time field is

145628-image.png

145667-image.png

and I would like to dynamically explicitly specify the sink data type mapping (not just the column name at sink ) . how can I do that?

Azure Data Factory
Azure Data Factory
An Azure service for ingesting, preparing, and transforming data at scale.
6,262 questions
{count} votes

1 answer

Sort by: Most helpful
  1. answered 2021-11-03T06:27:35.097+00:00
    ShaikMaheer-MSFT 28,096 Reputation points Microsoft Employee

    Hi @sakuraime ,

    Thank you for posting your Query.

    It seems, when we use Copy activity, we cannot control the data types of Sink. Its going to be automatically mapped by system.

    If we would like to have control on datatype of Sink too, then we can go with data flows.

    We can make use of derived column transformation in data flows, to convert datatype of any input column.
    For example, lets say we have ID column in source with type string and we want convert its type to int, then we can use expression toInteger(ID) inside derived column transformation