c21 commented on a change in pull request #29804: URL: https://github.com/apache/spark/pull/29804#discussion_r496854445
########## File path: sql/core/src/main/scala/org/apache/spark/sql/execution/bucketing/DisableUnnecessaryBucketedScan.scala ########## @@ -0,0 +1,153 @@ +/* + * 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. + */ + +package org.apache.spark.sql.execution.bucketing + +import org.apache.spark.sql.catalyst.plans.physical.{ClusteredDistribution, HashClusteredDistribution} +import org.apache.spark.sql.catalyst.rules.Rule +import org.apache.spark.sql.execution.{FileSourceScanExec, FilterExec, ProjectExec, SortExec, SparkPlan} +import org.apache.spark.sql.execution.aggregate.BaseAggregateExec +import org.apache.spark.sql.execution.exchange.Exchange +import org.apache.spark.sql.internal.SQLConf + +/** + * Disable unnecessary bucketed table scan based on actual physical query plan. + * NOTE: this rule is designed to be applied right after [[EnsureRequirements]], + * where all [[ShuffleExchangeExec]] and [[SortExec]] have been added to plan properly. + * + * When BUCKETING_ENABLED and AUTO_BUCKETED_SCAN_ENABLED are set to true, go through + * query plan to check where bucketed table scan is unnecessary, and disable bucketed table + * scan if: + * + * 1.The sub-plan from the nearest downstream [[hasInterestingPartition]] operator + * to the bucketed table scan, contains only [[isAllowedUnaryExecNode]] operators + * and at least one [[Exchange]]. + * + * 2.The sub-plan from root to bucketed table scan, does not contain + * [[hasInterestingPartition]] operator. + * + * Examples: + * 1.join: + * SortMergeJoin(t1.i = t2.j) + * / \ + * Sort(i) Sort(j) + * / \ + * Shuffle(i) Scan(t2: i, j) + * / (bucketed on column j, enable bucketed scan) + * Scan(t1: i, j) + * (bucketed on column j, DISABLE bucketed scan) + * + * 2.aggregate: + * HashAggregate(i, ..., Final) + * | + * Shuffle(i) + * | + * HashAggregate(i, ..., Partial) + * | + * Filter + * | + * Scan(t1: i, j) + * (bucketed on column j, DISABLE bucketed scan) Review comment: @cloud-fan - added. ########## File path: sql/core/src/main/scala/org/apache/spark/sql/execution/bucketing/DisableUnnecessaryBucketedScan.scala ########## @@ -0,0 +1,153 @@ +/* + * 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. + */ + +package org.apache.spark.sql.execution.bucketing + +import org.apache.spark.sql.catalyst.plans.physical.{ClusteredDistribution, HashClusteredDistribution} +import org.apache.spark.sql.catalyst.rules.Rule +import org.apache.spark.sql.execution.{FileSourceScanExec, FilterExec, ProjectExec, SortExec, SparkPlan} +import org.apache.spark.sql.execution.aggregate.BaseAggregateExec +import org.apache.spark.sql.execution.exchange.Exchange +import org.apache.spark.sql.internal.SQLConf + +/** + * Disable unnecessary bucketed table scan based on actual physical query plan. + * NOTE: this rule is designed to be applied right after [[EnsureRequirements]], + * where all [[ShuffleExchangeExec]] and [[SortExec]] have been added to plan properly. + * + * When BUCKETING_ENABLED and AUTO_BUCKETED_SCAN_ENABLED are set to true, go through + * query plan to check where bucketed table scan is unnecessary, and disable bucketed table + * scan if: + * + * 1.The sub-plan from the nearest downstream [[hasInterestingPartition]] operator + * to the bucketed table scan, contains only [[isAllowedUnaryExecNode]] operators + * and at least one [[Exchange]]. + * + * 2.The sub-plan from root to bucketed table scan, does not contain Review comment: @cloud-fan - updated. ########## File path: sql/core/src/test/scala/org/apache/spark/sql/sources/BucketedReadSuite.scala ########## @@ -1012,4 +1014,182 @@ abstract class BucketedReadSuite extends QueryTest with SQLTestUtils { } } } + + private def checkDisableBucketedScan( + query: String, Review comment: @cloud-fan - updated. ########## File path: sql/core/src/test/scala/org/apache/spark/sql/sources/BucketedReadSuite.scala ########## @@ -1012,4 +1014,182 @@ abstract class BucketedReadSuite extends QueryTest with SQLTestUtils { } } } + + private def checkDisableBucketedScan( + query: String, + expectedNumScanWithAutoScanEnabled: Int, + expectedNumScanWithAutoScanDisabled: Int): Unit = { + + def checkNumBucketedScan(query: String, expectedNumBucketedScan: Int): Unit = { + val plan = sql(query).queryExecution.executedPlan + val bucketedScan = plan.collect { case s: FileSourceScanExec if s.bucketedScan => s } + assert(bucketedScan.length == expectedNumBucketedScan) + } + + withSQLConf(SQLConf.AUTO_BUCKETED_SCAN_ENABLED.key -> "true") { + checkNumBucketedScan(query, expectedNumScanWithAutoScanEnabled) + val result = sql(query).collect() + + withSQLConf(SQLConf.AUTO_BUCKETED_SCAN_ENABLED.key -> "false") { + checkNumBucketedScan(query, expectedNumScanWithAutoScanDisabled) + checkAnswer(sql(query), result) + } + } + } + + test("SPARK-32859: disable unnecessary bucketed table scan - basic test") { Review comment: @cloud-fan - sure, moved to a separate test suite - `DisableUnnecessaryBucketedScanSuite`. ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
