NoneWriteDistributionSortedTableSpec: + See https://iceberg.apache.org/docs/1.6.0/spark-writes/#writing-distribution-modes Using write.distribution-mode - should create the appropriate number of Iceberg files + Given a table that has a distribution mode of none and is created with: CREATE TABLE polaris.my_namespace.NoneWriteDistributionSortedTableSpec ( id int, label String, partitionKey long, date Date, timestamp Timestamp ) USING iceberg TBLPROPERTIES ( 'format-version' = '2', 'write.spark.fanout.enabled' = 'true', 'write.distribution-mode' = 'none', 'sort-order' = 'partitionKey ASC NULLS FIRST' ) PARTITIONED BY (partitionKey); + And a query plan that looks like: == Optimized Logical Plan == Repartition 6, true, Statistics(sizeInBytes=6.3 MiB) +- Project [id#3975, partitionKey#3977L, date#3978, timestamp#3979, concat(cast(otherId#3993L as string), xxx) AS label#4012], Statistics(sizeInBytes=6.3 MiB) +- Join Inner, (partitionKey#3977L = otherId#3993L), Statistics(sizeInBytes=4.9 MiB) :- LocalRelation [id#3975, partitionKey#3977L, date#3978, timestamp#3979], Statistics(sizeInBytes=640.0 B, rowCount=20) +- Project [value#3990L AS otherId#3993L], Statistics(sizeInBytes=7.8 KiB) +- SerializeFromObject [input[0, bigint, false] AS value#3990L], Statistics(sizeInBytes=7.8 KiB) +- MapElements uk.co.odinconsultants.iceberg.distributions.AbstractWriteDistributionSpec$$Lambda$5586/0x0000000802566990@5c82f04, class java.lang.Long, [StructField(value,LongType,true)], obj#3989: bigint, Statistics(sizeInBytes=7.8 KiB) +- DeserializeToObject staticinvoke(class java.lang.Long, ObjectType(class java.lang.Long), valueOf, id#3985L, true, false, true), obj#3988: java.lang.Long, Statistics(sizeInBytes=7.8 KiB) +- Range (0, 1000, step=1, splits=Some(4)), Statistics(sizeInBytes=7.8 KiB, rowCount=1.00E+3) == Physical Plan == AdaptiveSparkPlan isFinalPlan=false +- Exchange RoundRobinPartitioning(6), REPARTITION_BY_NUM, [plan_id=6923] +- Project [id#3975, partitionKey#3977L, date#3978, timestamp#3979, concat(cast(otherId#3993L as string), xxx) AS label#4012] +- BroadcastHashJoin [partitionKey#3977L], [otherId#3993L], Inner, BuildLeft, false :- BroadcastExchange HashedRelationBroadcastMode(List(input[1, bigint, false]),false), [plan_id=6920] : +- LocalTableScan [id#3975, partitionKey#3977L, date#3978, timestamp#3979] +- Project [value#3990L AS otherId#3993L] +- SerializeFromObject [input[0, bigint, false] AS value#3990L] +- MapElements uk.co.odinconsultants.iceberg.distributions.AbstractWriteDistributionSpec$$Lambda$5586/0x0000000802566990@5c82f04, obj#3989: bigint +- DeserializeToObject staticinvoke(class java.lang.Long, ObjectType(class java.lang.Long), valueOf, id#3985L, true, false, true), obj#3988: java.lang.Long +- Range (0, 1000, step=1, splits=4) + And it has 20 rows over 30 data file(s) when writing with 4 executor threads + When we add another 20 rows of the same data that is logically distributed over 5 partition(s) + Then there are now 30 more data files + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +