pnowojski commented on a change in pull request #6787: [FLINK-8577][table] Implement proctime DataStream to Table upsert conversion URL: https://github.com/apache/flink/pull/6787#discussion_r249795484
########## File path: flink-libraries/flink-table/src/main/scala/org/apache/flink/table/plan/nodes/logical/FlinkLogicalUpsertToRetraction.scala ########## @@ -0,0 +1,74 @@ +/* + * 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.flink.table.plan.nodes.logical + +import java.util.{List => JList} + +import org.apache.calcite.plan._ +import org.apache.calcite.rel.{RelNode, SingleRel} +import org.apache.calcite.rel.convert.ConverterRule +import org.apache.calcite.rel.metadata.RelMetadataQuery +import org.apache.flink.table.plan.logical.rel.LogicalUpsertToRetraction +import org.apache.flink.table.plan.nodes.FlinkConventions + +class FlinkLogicalUpsertToRetraction( + cluster: RelOptCluster, + traitSet: RelTraitSet, + child: RelNode, + val keyNames: Seq[String]) + extends SingleRel(cluster, traitSet, child) + with FlinkLogicalRel { + + override def copy(traitSet: RelTraitSet, inputs: JList[RelNode]): RelNode = { + new FlinkLogicalUpsertToRetraction(cluster, traitSet, inputs.get(0), keyNames) + } + + override def computeSelfCost(planner: RelOptPlanner, mq: RelMetadataQuery): RelOptCost = { + val child = this.getInput + val rowCnt = mq.getRowCount(child) + // take rowCnt and fieldCnt into account, so that cost will be smaller when generate + // UpsertToRetractionConverter after Calc. + planner.getCostFactory.makeCost(rowCnt, rowCnt * child.getRowType.getFieldCount, 0) Review comment: Shouldn't you use `RelMetadataQuery#getAverageRowSize()` or at least duplicate the logic of `FlinkLogicalJoinBase#computeSelfCost` instead of using `getFieldCount`? Also Isn't the io cost missing? To be consistent with join cost this should return: ``` planner.getCostFactory.makeCost(rowCnt, rowCnt, rowCnt * estimateRowSize(child.getRowType)) ``` ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
