[
https://issues.apache.org/jira/browse/FLINK-7?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15041475#comment-15041475
]
ASF GitHub Bot commented on FLINK-7:
------------------------------------
Github user fhueske commented on a diff in the pull request:
https://github.com/apache/flink/pull/1255#discussion_r46674245
--- Diff:
flink-optimizer/src/main/java/org/apache/flink/optimizer/traversals/RangePartitionRewriter.java
---
@@ -0,0 +1,194 @@
+/*
+ * 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.optimizer.traversals;
+
+import org.apache.flink.api.common.distributions.CommonRangeBoundaries;
+import org.apache.flink.api.common.operators.UnaryOperatorInformation;
+import org.apache.flink.api.common.operators.base.GroupReduceOperatorBase;
+import org.apache.flink.api.common.operators.base.MapOperatorBase;
+import org.apache.flink.api.common.operators.base.MapPartitionOperatorBase;
+import org.apache.flink.api.common.operators.util.FieldList;
+import org.apache.flink.api.common.typeinfo.BasicTypeInfo;
+import org.apache.flink.api.common.typeinfo.TypeInformation;
+import org.apache.flink.api.common.typeutils.TypeComparatorFactory;
+import org.apache.flink.api.java.typeutils.RecordTypeInfo;
+import org.apache.flink.runtime.io.network.DataExchangeMode;
+import org.apache.flink.runtime.operators.udf.AssignRangeIndex;
+import org.apache.flink.runtime.operators.udf.PartitionIDRemoveWrapper;
+import org.apache.flink.runtime.operators.udf.RangeBoundaryBuilder;
+import org.apache.flink.api.java.functions.SampleInCoordinator;
+import org.apache.flink.api.java.functions.SampleInPartition;
+import org.apache.flink.api.java.sampling.IntermediateSampleData;
+import org.apache.flink.api.java.tuple.Tuple2;
+import org.apache.flink.api.java.typeutils.TupleTypeInfo;
+import org.apache.flink.api.java.typeutils.TypeExtractor;
+import org.apache.flink.optimizer.dag.GroupReduceNode;
+import org.apache.flink.optimizer.dag.MapNode;
+import org.apache.flink.optimizer.dag.MapPartitionNode;
+import org.apache.flink.optimizer.dag.TempMode;
+import org.apache.flink.optimizer.plan.Channel;
+import org.apache.flink.optimizer.plan.NamedChannel;
+import org.apache.flink.optimizer.plan.OptimizedPlan;
+import org.apache.flink.optimizer.plan.PlanNode;
+import org.apache.flink.optimizer.plan.SingleInputPlanNode;
+import org.apache.flink.optimizer.util.Utils;
+import org.apache.flink.runtime.operators.DriverStrategy;
+import org.apache.flink.runtime.operators.shipping.ShipStrategyType;
+import org.apache.flink.util.Visitor;
+
+import java.util.ArrayList;
+import java.util.LinkedList;
+import java.util.List;
+
+public class RangePartitionRewriter implements Visitor<PlanNode> {
+ final static long SEED = org.apache.flink.api.java.Utils.RNG.nextLong();
+
+ final OptimizedPlan plan;
+
+ public RangePartitionRewriter(OptimizedPlan plan) {
+ this.plan = plan;
+ }
+
+ @Override
+ public boolean preVisit(PlanNode visitable) {
+ return true;
+ }
+
+ @Override
+ public void postVisit(PlanNode visitable) {
+ final List<Channel> outgoingChannels =
visitable.getOutgoingChannels();
+ final List<Channel> newOutGoingChannels = new LinkedList<>();
+ final List<Channel> toBeRemoveChannels = new ArrayList<>();
+ for (Channel channel : outgoingChannels) {
+ ShipStrategyType shipStrategy =
channel.getShipStrategy();
+ if (shipStrategy == ShipStrategyType.PARTITION_RANGE) {
+ TypeInformation<?> outputType =
channel.getSource().getProgramOperator().getOperatorInfo().getOutputType();
+ // Do not optimize for record type, it's a
special case for range partitioner, and should be removed later.
+ if (!(outputType instanceof RecordTypeInfo)) {
--- End diff --
I think the `RecordOutputEmitter` would fail to partition by range because
it does not have access to the generated data distribution. You are right, it
would not silently produce wrong results, but probably fail at runtime with a
`NullPointerException` in line 193 of `RecordOutputEmitter`.
Good observation with the `GenericDataSinkBase`! The methods that relate to
range partitioning and global order are artifacts from the recently removed
Record API and not publicly accessible. In fact, they are useless, because they
lack the definition of a key on which the output should be partitioned /
ordered.
I will open a PR to remove the `setGlobalOrder()`, `setRangePartitioned()`,
`getDataDistribution()`, and `getPartitioningOrder()` methods from the
`GenericDataSinkBase`.
> [GitHub] Enable Range Partitioner
> ---------------------------------
>
> Key: FLINK-7
> URL: https://issues.apache.org/jira/browse/FLINK-7
> Project: Flink
> Issue Type: Sub-task
> Components: Distributed Runtime
> Reporter: GitHub Import
> Assignee: Chengxiang Li
> Fix For: pre-apache
>
>
> The range partitioner is currently disabled. We need to implement the
> following aspects:
> 1) Distribution information, if available, must be propagated back together
> with the ordering property.
> 2) A generic bucket lookup structure (currently specific to PactRecord).
> Tests to re-enable after fixing this issue:
> - TeraSortITCase
> - GlobalSortingITCase
> - GlobalSortingMixedOrderITCase
> ---------------- Imported from GitHub ----------------
> Url: https://github.com/stratosphere/stratosphere/issues/7
> Created by: [StephanEwen|https://github.com/StephanEwen]
> Labels: core, enhancement, optimizer,
> Milestone: Release 0.4
> Assignee: [fhueske|https://github.com/fhueske]
> Created at: Fri Apr 26 13:48:24 CEST 2013
> State: open
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
