Hi, We would like to be able to use the iceberg spark Datasource (IcebergSource) to write kafka sourced streaming dataframes.
In tests, we are able to succesfully create a partitioned table and write when using the MemoryStream, but when using a kafka source: *spark.readStream.format("kafka")* and writing to iceberg: *dataFrame.writeStream .format("catwalk-iceberg") .outputMode(OutputMode.Append) .trigger(Trigger.Once) .option("path", uri.toString) .option("checkpointLocation", Paths.get(uri.toString, "checkpoint").toString) .start .awaitTermination* we get this exception: Caused by: java.lang.IllegalStateException: Already closed file for partition: happened_at_day=2000-01-01 at org.apache.iceberg.spark.source.Writer$PartitionedWriter.write(Writer.java:389) at org.apache.iceberg.spark.source.Writer$PartitionedWriter.write(Writer.java:350) Before I dig deeper, is this something that has worked for anyone? Thanks!