fvaleye commented on code in PR #1620: URL: https://github.com/apache/iceberg-rust/pull/1620#discussion_r2313751052
########## crates/integrations/datafusion/src/physical_plan/repartition.rs: ########## @@ -0,0 +1,805 @@ +// Licensed to the Apache Software Foundation (ASF) under one +// or more contributor license agreements. See the NOTICE file +// distributed with this work for additional information +// regarding copyright ownership. The ASF licenses this file +// to you under the Apache License, Version 2.0 (the +// "License"); you may not use this file except in compliance +// with the License. You may obtain a copy of the License at +// +// http://www.apache.org/licenses/LICENSE-2.0 +// +// Unless required by applicable law or agreed to in writing, +// software distributed under the License is distributed on an +// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +// KIND, either express or implied. See the License for the +// specific language governing permissions and limitations +// under the License. + +use std::collections::HashSet; +use std::sync::Arc; + +use datafusion::error::Result as DFResult; +use datafusion::physical_plan::expressions::Column; +use datafusion::physical_plan::repartition::RepartitionExec; +use datafusion::physical_plan::{ExecutionPlan, Partitioning}; +use iceberg::spec::{SchemaRef, TableMetadata, TableMetadataRef, Transform}; + +/// Creates an Iceberg-aware repartition execution plan that optimizes data distribution +/// for parallel processing while respecting Iceberg table partitioning semantics. +/// +/// This function automatically determines the optimal partitioning strategy based on +/// the table's partition specification and sort order: +/// +/// ## Partitioning Strategies +/// +/// - **Unpartitioned tables**: Uses round-robin distribution to ensure balanced load +/// across all workers, maximizing parallelism for write operations. +/// +/// - **Hash partitioning**: Used for tables with identity transforms or bucket transforms: +/// - Identity partition columns (e.g., `PARTITIONED BY (user_id, category)`) +/// - Bucket columns from partition spec or sort order +/// - This ensures data co-location within partitions and buckets for optimal file clustering +/// +/// - **Round-robin partitioning**: Used for: +/// - Range-only partitions (e.g., date/time partitions that concentrate data) +/// - Tables with only temporal/range transforms that don't provide good distribution +/// - Unpartitioned or non-bucketed tables +/// +/// - **Mixed transforms**: Tables with both range and identity/bucket transforms use hash +/// partitioning on the identity/bucket columns for optimal distribution. +/// +/// ## Performance notes +/// +/// - Only repartitions when the input partitioning scheme differs from the desired strategy +/// - Only repartitions when the input partition count differs from the target +/// - Requires explicit target partition count for deterministic behavior +/// - Preserves column order (partitions first, then buckets) for consistent file layout +/// +/// # Arguments +/// +/// * `input` - The input execution plan providing data to be repartitioned +/// * `table_schema` - The Iceberg table schema used to resolve column references +/// * `table_metadata` - The Iceberg table metadata containing partition spec and sort order +/// * `target_partitions` - Target number of partitions for parallel processing (must be > 0) +/// +/// # Returns +/// +/// An execution plan that will apply the optimal partitioning strategy during execution, +/// or the original input plan unchanged if no repartitioning is needed. +/// +/// # Example +/// +/// ```ignore +/// let repartitioned_plan = repartition( +/// input_plan, +/// table.schema_ref(), +/// table.metadata_ref(), +/// 4, // Explicit partition count +/// )?; +/// ``` +pub fn repartition( Review Comment: True, it will be part of the `ExecutionPlan` building. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org For additional commands, e-mail: issues-h...@iceberg.apache.org
