zentol commented on a change in pull request #10378: [FLINK-14919] [flink-end-to-end-perf-tests] Add performance e2e test suite for basic operations URL: https://github.com/apache/flink/pull/10378#discussion_r366981360
########## File path: flink-end-to-end-perf-tests/flink-basic-operations/src/main/java/org/apache/flink/basic/operations/PerformanceTestJob.java ########## @@ -0,0 +1,225 @@ +/* + * 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.basic.operations; + +import org.apache.flink.api.common.ExecutionMode; +import org.apache.flink.api.common.typeinfo.BasicTypeInfo; +import org.apache.flink.api.common.typeinfo.Types; +import org.apache.flink.api.java.tuple.Tuple2; +import org.apache.flink.api.java.tuple.Tuple3; +import org.apache.flink.api.java.typeutils.TupleTypeInfo; +import org.apache.flink.api.java.utils.ParameterTool; +import org.apache.flink.basic.utils.FromElements; +import org.apache.flink.basic.utils.ValueStateFlatMap; +import org.apache.flink.basic.utils.WordCountData; +import org.apache.flink.contrib.streaming.state.RocksDBStateBackend; +import org.apache.flink.core.fs.FileSystem.WriteMode; +import org.apache.flink.core.fs.Path; +import org.apache.flink.runtime.jobgraph.ScheduleMode; +import org.apache.flink.runtime.state.filesystem.FsStateBackend; +import org.apache.flink.streaming.api.CheckpointingMode; +import org.apache.flink.streaming.api.TimeCharacteristic; +import org.apache.flink.streaming.api.datastream.DataStream; +import org.apache.flink.streaming.api.environment.StreamExecutionEnvironment; +import org.apache.flink.streaming.api.functions.co.CoProcessFunction; +import org.apache.flink.streaming.api.graph.StreamGraph; +import org.apache.flink.util.Collector; + +/** + * End-to-end perf test for end-to-end perf test. + * + * <p>The sources are generated and bounded. The result is always constant. + * + * <p>Parameters: + * --jobName job's name + * --parallelism source Node's parallelism + * --checkPointMode checkpointMode，atleastone or exactlyonce + * --seed + * --sleepNum + * --streamPartitioner + * --scheduleMode + * --executionMode + * --inputPath if source data in hdfs then input the inputPath + * --checkpointInterval + * --checkpointPath + * --stateBackend + * --checkpointTimeout + */ +public class PerformanceTestJob { + private static byte[] size10B = new byte[10]; + private static String recordSize10B = new String(size10B); + private static byte[] size100B = new byte[100]; + private static String recordSize100B = new String(size100B); + private static byte[] size1KB = new byte[1024]; + private static String recordSize1KB = new String(size1KB); + private static String recordValue; + + protected DataStream<Tuple2<String, String>> generateGraph( + String sourceName, + StreamExecutionEnvironment env, + int recordSize, + long maxCount, + int sleepNum, + int seed, + int parallism, + String streamPartitioner, + String outputPath){ + if (recordSize == 10){ + recordValue = recordSize10B; + } else if (recordSize == 100){ + recordValue = recordSize100B; + } else { + recordValue = recordSize1KB; + } + + DataStream<String> sourceNode = FromElements.fromElements(env, maxCount, sleepNum, + WordCountData.getWords(recordValue.length(), true, seed)).setParallelism(parallism).name(sourceName); + + DataStream<Tuple2<String, String>> flapNode = + sourceNode + .flatMap((String value, Collector<Tuple2<String, String>> collector) + -> collector.collect(new Tuple2<String, String>(value, recordValue))) + .returns(Types.TUPLE(Types.STRING, Types.STRING)) + .setParallelism(parallism) + .keyBy(0) + .flatMap(new ValueStateFlatMap()); + + DataStream<Tuple2<String, String>> outputNode = null; + + if ("forward".equals(streamPartitioner)) { + outputNode = flapNode.forward(); + } else if ("rescale".equals(streamPartitioner)) { + outputNode = flapNode.rescale(); + } else if ("rebalance".equals(streamPartitioner)) { + outputNode = flapNode.rebalance(); + } else if ("shuffle".equals(streamPartitioner)) { + outputNode = flapNode.shuffle(); + } else if ("keyby".equals(streamPartitioner)) { + outputNode = flapNode.keyBy(0); + } else if ("broadcast".equals(streamPartitioner)) { + outputNode = flapNode.broadcast(); + } else { + throw new IllegalArgumentException("Argument streamPartitioner is illegal!"); + } + return outputNode; + + } + + protected DataStream<Tuple3<String, String, String>> connector( + DataStream<Tuple2<String, String>> output1, + DataStream<Tuple2<String, String>> output2, + int parallelism){ + DataStream<Tuple3<String, String, String>> output = output1.connect(output2).process( + new CoProcessFunction<Tuple2<String, String>, Tuple2<String, String>, Tuple3<String, String, String>>() { + @Override + public void processElement1(Tuple2<String, String> value, Context ctx, Collector<Tuple3<String, String, String>> out) throws Exception { + out.collect(new Tuple3<>(value.f0, value.f1, value.f1)); + } + + @Override + public void processElement2(Tuple2<String, String> value, Context ctx, Collector<Tuple3<String, String, String>>out) throws Exception { + processElement1(value, ctx, out); + } + + }, new TupleTypeInfo<>(BasicTypeInfo.STRING_TYPE_INFO, BasicTypeInfo.STRING_TYPE_INFO, Review comment: why do we have to explicitly pass the return type? ---------------------------------------------------------------- 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: us...@infra.apache.org With regards, Apache Git Services
