S3OutputFormatConfig - AmazonAppFlow

S3OutputFormatConfig

The configuration that determines how Amazon AppFlow should format the flow output data when Amazon S3 is used as the destination.

Contents

aggregationConfig

The aggregation settings that you can use to customize the output format of your flow data.

Type: AggregationConfig object

Required: No

fileType

Indicates the file type that Amazon AppFlow places in the Amazon S3 bucket.

Type: String

Valid Values: CSV | JSON | PARQUET

Required: No

prefixConfig

Determines the prefix that Amazon AppFlow applies to the folder name in the Amazon S3 bucket. You can name folders according to the flow frequency and date.

Type: PrefixConfig object

Required: No

preserveSourceDataTyping

If your file output format is Parquet, use this parameter to set whether Amazon AppFlow preserves the data types in your source data when it writes the output to Amazon S3.

  • true: Amazon AppFlow preserves the data types when it writes to Amazon S3. For example, an integer or 1 in your source data is still an integer in your output.

  • false: Amazon AppFlow converts all of the source data into strings when it writes to Amazon S3. For example, an integer of 1 in your source data becomes the string "1" in the output.

Type: Boolean

Required: No

See Also

For more information about using this API in one of the language-specific AWS SDKs, see the following: