pvary commented on code in PR #8553: URL: https://github.com/apache/iceberg/pull/8553#discussion_r1401866044
########## flink/v1.17/flink/src/test/java/org/apache/iceberg/flink/source/TestIcebergSourceWithWatermarkExtractor.java: ########## @@ -0,0 +1,451 @@ +/* + * 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.iceberg.flink.source; + +import static org.apache.flink.connector.testframe.utils.ConnectorTestConstants.DEFAULT_COLLECT_DATA_TIMEOUT; +import static org.assertj.core.api.AssertionsForClassTypes.assertThat; + +import java.io.Serializable; +import java.time.Duration; +import java.time.Instant; +import java.time.LocalDateTime; +import java.time.ZoneId; +import java.util.Collection; +import java.util.List; +import java.util.Optional; +import java.util.Set; +import java.util.concurrent.CompletableFuture; +import java.util.concurrent.ConcurrentMap; +import java.util.concurrent.TimeUnit; +import java.util.concurrent.atomic.AtomicInteger; +import java.util.stream.Collectors; +import org.apache.flink.api.common.JobID; +import org.apache.flink.api.common.eventtime.SerializableTimestampAssigner; +import org.apache.flink.api.common.eventtime.WatermarkStrategy; +import org.apache.flink.api.common.typeinfo.TypeInformation; +import org.apache.flink.configuration.Configuration; +import org.apache.flink.core.execution.JobClient; +import org.apache.flink.metrics.Gauge; +import org.apache.flink.runtime.metrics.MetricNames; +import org.apache.flink.runtime.minicluster.RpcServiceSharing; +import org.apache.flink.runtime.testutils.InMemoryReporter; +import org.apache.flink.runtime.testutils.MiniClusterResourceConfiguration; +import org.apache.flink.streaming.api.datastream.DataStream; +import org.apache.flink.streaming.api.environment.StreamExecutionEnvironment; +import org.apache.flink.streaming.api.functions.windowing.AllWindowFunction; +import org.apache.flink.streaming.api.windowing.assigners.TumblingEventTimeWindows; +import org.apache.flink.streaming.api.windowing.time.Time; +import org.apache.flink.streaming.api.windowing.windows.TimeWindow; +import org.apache.flink.table.data.GenericRowData; +import org.apache.flink.table.data.RowData; +import org.apache.flink.test.util.MiniClusterWithClientResource; +import org.apache.flink.util.CloseableIterator; +import org.apache.flink.util.Collector; +import org.apache.iceberg.FileFormat; +import org.apache.iceberg.data.GenericAppenderHelper; +import org.apache.iceberg.data.GenericRecord; +import org.apache.iceberg.data.Record; +import org.apache.iceberg.flink.HadoopTableResource; +import org.apache.iceberg.flink.RowDataConverter; +import org.apache.iceberg.flink.TestFixtures; +import org.apache.iceberg.relocated.com.google.common.collect.ImmutableList; +import org.apache.iceberg.relocated.com.google.common.collect.ImmutableMap; +import org.apache.iceberg.relocated.com.google.common.collect.Lists; +import org.apache.iceberg.relocated.com.google.common.collect.Maps; +import org.apache.iceberg.relocated.com.google.common.collect.Sets; +import org.awaitility.Awaitility; +import org.junit.Assert; +import org.junit.ClassRule; +import org.junit.Rule; +import org.junit.Test; +import org.junit.rules.TemporaryFolder; + +public class TestIcebergSourceWithWatermarkExtractor implements Serializable { + private static final InMemoryReporter reporter = InMemoryReporter.createWithRetainedMetrics(); + private static final int PARALLELISM = 4; + private static final String SOURCE_NAME = "IcebergSource"; + private static final int RECORD_NUM_FOR_2_SPLITS = 200; + private static final ConcurrentMap<Long, Integer> windows = Maps.newConcurrentMap(); + + @ClassRule public static final TemporaryFolder TEMPORARY_FOLDER = new TemporaryFolder(); + + @Rule + public final MiniClusterWithClientResource miniClusterResource = + new MiniClusterWithClientResource( + new MiniClusterResourceConfiguration.Builder() + .setNumberTaskManagers(1) + .setNumberSlotsPerTaskManager(PARALLELISM) + .setRpcServiceSharing(RpcServiceSharing.DEDICATED) + .setConfiguration(reporter.addToConfiguration(new Configuration())) + .withHaLeadershipControl() + .build()); + + @Rule + public final HadoopTableResource sourceTableResource = + new HadoopTableResource( + TEMPORARY_FOLDER, TestFixtures.DATABASE, TestFixtures.TABLE, TestFixtures.TS_SCHEMA); + + /** + * This is an integration test for watermark handling and windowing. Integration testing the + * following features: + * + * <ul> + * <li>- Ordering of the splits + * <li>- Emitting of watermarks + * <li>- Firing windows based on watermarks + * </ul> + * + * <p>The test generates 4 splits + * + * <ul> + * <li>- Split 1 - Watermark 100 min + * <li>- Split 2, 3 - Watermark 0 min + * <li>- Split 4 - Watermark 6 min + * </ul> + * + * <p>Creates a source with 5 minutes tumbling window with parallelism 1 (to prevent concurrency + * issues). + * + * <p>Checks that windows are handled correctly based on the emitted watermarks, and splits are + * read in the following order: + * + * <ul> + * <li>- Split 2, 3 + * <li>- Split 4 + * <li>- Split 1 + * </ul> + * + * <p>As a result the window aggregator emits the records based on in Split 2-3, and Split 4 data. + * + * <p>Add 2 more splits, so the task manager close the windows for the original 4 splits and emit + * the appropriate aggregated records. + */ + @Test + public void testWindowing() throws Exception { + GenericAppenderHelper dataAppender = appender(); + List<Record> expectedRecords = Lists.newArrayList(); + + // Generate records with the following pattern: + // - File 1 - Later records (Watermark 6000000) + // - Split 1 - 2 records (100, "file_1-recordTs_100"), (103, "file_1-recordTs_103") + // - File 2 - First records (Watermark 0) + // - Split 1 - 100 records (0, "file_2-recordTs_0"), (1, "file_2-recordTs_1"),... + // - Split 2 - 100 records (0, "file_2-recordTs_0"), (1, "file_2-recordTs_1"),... + // - File 3 - Parallel write for the first records (Watermark 360000) + // - Split 1 - 2 records (6, "file_3-recordTs_6"), (7, "file_3-recordTs_7") + List<Record> batch = + ImmutableList.of( + generateRecord(100, "file_1-recordTs_100"), + generateRecord(101, "file_1-recordTs_101"), + generateRecord(103, "file_1-recordTs_103")); + expectedRecords.addAll(batch); + dataAppender.appendToTable(batch); + + batch = Lists.newArrayListWithCapacity(100); + for (int i = 0; i < RECORD_NUM_FOR_2_SPLITS; ++i) { + // Generate records where the timestamps are out of order, but still between 0-5 minutes + batch.add(generateRecord(4 - i % 5, "file_2-recordTs_" + i)); + } + expectedRecords.addAll(batch); + dataAppender.appendToTable(batch); + + batch = + ImmutableList.of( + generateRecord(6, "file_3-recordTs_6"), generateRecord(7, "file_3-recordTs_7")); + expectedRecords.addAll(batch); + dataAppender.appendToTable(batch); + + StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment(); + env.setParallelism(1); + + DataStream<RowData> stream = + env.fromSource( + source(), + WatermarkStrategy.<RowData>noWatermarks() + .withTimestampAssigner(new RowDataTimestampAssigner()), + SOURCE_NAME, + TypeInformation.of(RowData.class)); + + stream + .windowAll(TumblingEventTimeWindows.of(Time.minutes(5))) + .apply( + new AllWindowFunction<RowData, RowData, TimeWindow>() { + @Override + public void apply( + TimeWindow window, Iterable<RowData> values, Collector<RowData> out) { + // Emit RowData which contains the window start time, and the record count in + // that window + AtomicInteger count = new AtomicInteger(0); + values.forEach(a -> count.incrementAndGet()); + out.collect(row(window.getStart(), count.get())); + windows.put(window.getStart(), count.get()); + } + }); + + // Use static variable to collect the windows, since other solutions were flaky + windows.clear(); Review Comment: This is somewhat inelegant, but the `windowed.collectAsync()` solution was flaky. Failed once per ~1000 runs. Even though the window operator was called, and emitted the records (checked by logs), the results did not show up in the `resultIterator`. Maybe some Flink glitch? So opted for a static `ConcurrentMap` instead to collect the generated windows. -- 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. To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org For additional commands, e-mail: issues-h...@iceberg.apache.org
