Hi all, I am currently working with flink reading from a kinesis stream and writing to S3. In Kinesis firehose you are able to dynamically partition the streaming data into prefixes in s3 based on keys within the data. I am wondering if flink provides similar features with a DataStream FileSink? Thanks.

Check BucketAssigner in Flink https://nightlies.apache.org/flink/flink-docs-release-1.17/docs/connectors/datastream/filesystem/#bucket-assignment You can also extend the BucketAssigner and create your own