Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

TEZ-3268: Provide a demuxer sample app that uses fair routing #320

Open
wants to merge 2 commits into
base: master
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from 1 commit
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
231 changes: 231 additions & 0 deletions tez-examples/src/main/java/org/apache/tez/examples/Demuxer.java
Original file line number Diff line number Diff line change
@@ -0,0 +1,231 @@
/**
* 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.tez.examples;

import java.util.Arrays;
import java.util.List;
import java.util.stream.Collectors;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.lib.input.TextInputFormat;
import org.apache.hadoop.mapreduce.lib.output.TextOutputFormat;
import org.apache.hadoop.util.ToolRunner;
import org.apache.tez.client.TezClient;
import org.apache.tez.common.Preconditions;
import org.apache.tez.dag.api.DAG;
import org.apache.tez.dag.api.Edge;
import org.apache.tez.dag.api.GroupInputEdge;
import org.apache.tez.dag.api.InputDescriptor;
import org.apache.tez.dag.api.ProcessorDescriptor;
import org.apache.tez.dag.api.TezConfiguration;
import org.apache.tez.dag.api.Vertex;
import org.apache.tez.dag.library.vertexmanager.FairShuffleVertexManager;
import org.apache.tez.dag.library.vertexmanager.FairShuffleVertexManager.FairRoutingType;
import org.apache.tez.examples.HashJoinExample.ForwardingProcessor;
import org.apache.tez.mapreduce.input.MRInput;
import org.apache.tez.mapreduce.output.MultiMROutput;
import org.apache.tez.mapreduce.processor.SimpleMRProcessor;
import org.apache.tez.runtime.api.LogicalInput;
import org.apache.tez.runtime.api.ProcessorContext;
import org.apache.tez.runtime.library.api.KeyValueWriterWithBasePath;
import org.apache.tez.runtime.library.api.KeyValuesReader;
import org.apache.tez.runtime.library.api.TezRuntimeConfiguration;
import org.apache.tez.runtime.library.api.TezRuntimeConfiguration.ReportPartitionStats;
import org.apache.tez.runtime.library.conf.OrderedPartitionedKVEdgeConfig;
import org.apache.tez.runtime.library.input.ConcatenatedMergedKeyValuesInput;
import org.apache.tez.runtime.library.partitioner.HashPartitioner;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

/**
* Simple example of a demuxer tolerant of data skew problems. The example scans the source files,
* partitions records by values, and write them in /path/to/{value}-r-{task id}.
*/
public class Demuxer extends TezExampleBase {
private static final Logger LOG = LoggerFactory.getLogger(Demuxer.class);

private static final String DEMUXER_OUTPUT = "demuxerOutput";

public static void main(String[] args) throws Exception {
Demuxer job = new Demuxer();
int status = ToolRunner.run(new Configuration(), job, args);
System.exit(status);
}

@Override
protected void printUsage() {
System.err.println("Usage: "
+ "demuxer <inputPath> <outPath> <numPartitions> [isPrecise(default false)]");
}

@Override
protected int validateArgs(String[] otherArgs) {
if (otherArgs.length < 3 || otherArgs.length > 4) {
return 2;
}
return 0;
}

@Override
protected int runJob(String[] args, TezConfiguration tezConf, TezClient tezClient)
throws Exception {
String inputDirs = args[0];
String outputDir = args[1];
int numPartitions = Integer.parseInt(args[2]);
if (args.length == 4 && Boolean.parseBoolean(args[3])) {
tezConf.set(
TezRuntimeConfiguration.TEZ_RUNTIME_REPORT_PARTITION_STATS,
ReportPartitionStats.PRECISE.getType());
}

List<Path> inputPaths = Arrays
.stream(inputDirs.split(","))
.map(Path::new)
.collect(Collectors.toList());
Path outputPath = new Path(outputDir);

FileSystem fs = outputPath.getFileSystem(tezConf);
outputPath = fs.makeQualified(outputPath);
if (fs.exists(outputPath)) {
System.err.println("Output directory: " + outputDir + " already exists");
return 3;
}
DAG dag = inputPaths.size() == 1
? createDag(tezConf, inputPaths.get(0), outputPath, numPartitions)
: createDagWithUnion(tezConf, inputPaths, outputPath, numPartitions);
LOG.info("Running Demuxer");
return runDag(dag, isCountersLog(), LOG);
}

private DAG createDag(TezConfiguration tezConf, Path inputPath, Path outputPath,
int numPartitions) {
Vertex inputVertex = createInputVertex(tezConf, "input", inputPath);
Vertex demuxVertex = createDemuxVertex(tezConf, outputPath, numPartitions);
OrderedPartitionedKVEdgeConfig edgeConf = OrderedPartitionedKVEdgeConfig
.newBuilder(
Text.class.getName(),
NullWritable.class.getName(),
HashPartitioner.class.getName())
.setFromConfiguration(tezConf)
.build();
return DAG
.create("Demuxer")
.addVertex(inputVertex)
.addVertex(demuxVertex)
.addEdge(Edge.create(inputVertex, demuxVertex, edgeConf.createDefaultEdgeProperty()));
}

private DAG createDagWithUnion(TezConfiguration tezConf, List<Path> inputPaths, Path outputPath,
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Note: This case doesn't work yet until we merge #306

Copy link
Contributor

@abstractdog abstractdog Aug 4, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

#306 was closed due to some reasons
is this PR suitable to be adapted to work without that patch? if so, still interested in seeing this in the codebase

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@abstractdog Thanks for asking. I closed #306 and still need to implement an alternative solution. At this point, I removed the case using UNION from the sample.
9e09665

int numPartitions) {
Vertex[] inputVertices = new Vertex[inputPaths.size()];
for (int i = 0; i < inputPaths.size(); i++) {
inputVertices[i] = createInputVertex(tezConf, "input-" + i, inputPaths.get(i));
}

Vertex demuxVertex = createDemuxVertex(tezConf, outputPath, numPartitions);

OrderedPartitionedKVEdgeConfig edgeConf = OrderedPartitionedKVEdgeConfig
.newBuilder(Text.class.getName(), NullWritable.class.getName(),
HashPartitioner.class.getName())
.setFromConfiguration(tezConf)
.build();

DAG dag = DAG.create("Demuxer");
Arrays.stream(inputVertices).forEach(dag::addVertex);
return dag
.addVertex(demuxVertex)
.addEdge(
GroupInputEdge.create(
dag.createVertexGroup("union", inputVertices),
demuxVertex,
edgeConf.createDefaultEdgeProperty(),
InputDescriptor.create(ConcatenatedMergedKeyValuesInput.class.getName())));
}

private Vertex createInputVertex(TezConfiguration tezConf, String vertexName, Path path) {
// This vertex represents an input vertex for the demuxer. It reads text data using the
// TextInputFormat. ForwardingProcessor simply forwards the data downstream as is.
return Vertex
.create(vertexName, ProcessorDescriptor.create(ForwardingProcessor.class.getName()))
.addDataSource(
"inputFile",
MRInput
.createConfigBuilder(
new Configuration(tezConf),
TextInputFormat.class,
path.toUri().toString())
.groupSplits(!isDisableSplitGrouping())
.generateSplitsInAM(!isGenerateSplitInClient()).build());
}

private Vertex createDemuxVertex(TezConfiguration tezConf, Path outputPath, int numPartitions) {
// This vertex demuxes records based on the keys. Multiple reduce tasks can process the same key
// as fair routing is configured.
return Vertex
.create(
"demuxer",
ProcessorDescriptor.create(DemuxProcessor.class.getName()), numPartitions)
.setVertexManagerPlugin(
FairShuffleVertexManager
.createConfigBuilder(tezConf)
.setAutoParallelism(FairRoutingType.FAIR_PARALLELISM)
// These params demonstrate perfect fair routing
.setSlowStartMinSrcCompletionFraction(1.0f)
.setSlowStartMaxSrcCompletionFraction(1.0f)
.build())
.addDataSink(
DEMUXER_OUTPUT,
MultiMROutput
.createConfigBuilder(
new Configuration(tezConf),
TextOutputFormat.class,
outputPath.toUri().toString(),
false)
.build());
}

public static class DemuxProcessor extends SimpleMRProcessor {
public DemuxProcessor(ProcessorContext context) {
super(context);
}

@Override
public void run() throws Exception {
Preconditions.checkArgument(!getInputs().isEmpty());
Preconditions.checkArgument(getOutputs().size() == 1);
KeyValueWriterWithBasePath kvWriter = (KeyValueWriterWithBasePath) getOutputs()
.get(DEMUXER_OUTPUT)
.getWriter();
for (LogicalInput input : getInputs().values()) {
KeyValuesReader kvReader = (KeyValuesReader) input.getReader();
while (kvReader.next()) {
Text category = (Text) kvReader.getCurrentKey();
String path = category.toString();
for (Object value : kvReader.getCurrentValues()) {
assert value == NullWritable.get();
kvWriter.write(category, NullWritable.get(), path);
}
}
}
}
}
}
Loading
Loading