Search in sources :

Example 1 with IndexType

use of org.apache.hudi.index.HoodieIndex.IndexType in project hudi by apache.

the class TestHoodieIndex method testSimpleTagLocationAndUpdate.

@ParameterizedTest
@MethodSource("indexTypeParams")
public void testSimpleTagLocationAndUpdate(IndexType indexType, boolean populateMetaFields, boolean enableMetadataIndex) throws Exception {
    setUp(indexType, populateMetaFields, enableMetadataIndex);
    String newCommitTime = "001";
    int totalRecords = 10 + random.nextInt(20);
    List<HoodieRecord> records = dataGen.generateInserts(newCommitTime, totalRecords);
    JavaRDD<HoodieRecord> writeRecords = jsc.parallelize(records, 1);
    metaClient = HoodieTableMetaClient.reload(metaClient);
    HoodieTable hoodieTable = HoodieSparkTable.create(config, context, metaClient);
    // Test tagLocation without any entries in index
    JavaRDD<HoodieRecord> javaRDD = tagLocation(index, writeRecords, hoodieTable);
    assert (javaRDD.filter(record -> record.isCurrentLocationKnown()).collect().size() == 0);
    // Insert totalRecords records
    writeClient.startCommitWithTime(newCommitTime);
    JavaRDD<WriteStatus> writeStatues = writeClient.upsert(writeRecords, newCommitTime);
    Assertions.assertNoWriteErrors(writeStatues.collect());
    // Now tagLocation for these records, index should not tag them since it was a failed
    // commit
    javaRDD = tagLocation(index, writeRecords, hoodieTable);
    assert (javaRDD.filter(record -> record.isCurrentLocationKnown()).collect().size() == 0);
    // Now commit this & update location of records inserted and validate no errors
    writeClient.commit(newCommitTime, writeStatues);
    // Now tagLocation for these records, index should tag them correctly
    metaClient = HoodieTableMetaClient.reload(metaClient);
    hoodieTable = HoodieSparkTable.create(config, context, metaClient);
    javaRDD = tagLocation(index, writeRecords, hoodieTable);
    Map<String, String> recordKeyToPartitionPathMap = new HashMap();
    List<HoodieRecord> hoodieRecords = writeRecords.collect();
    hoodieRecords.forEach(entry -> recordKeyToPartitionPathMap.put(entry.getRecordKey(), entry.getPartitionPath()));
    assertEquals(totalRecords, javaRDD.filter(record -> record.isCurrentLocationKnown()).collect().size());
    assertEquals(totalRecords, javaRDD.map(record -> record.getKey().getRecordKey()).distinct().count());
    assertEquals(totalRecords, javaRDD.filter(record -> (record.getCurrentLocation() != null && record.getCurrentLocation().getInstantTime().equals(newCommitTime))).distinct().count());
    javaRDD.foreach(entry -> assertEquals(recordKeyToPartitionPathMap.get(entry.getRecordKey()), entry.getPartitionPath(), "PartitionPath mismatch"));
    JavaRDD<HoodieKey> hoodieKeyJavaRDD = writeRecords.map(entry -> entry.getKey());
    JavaPairRDD<HoodieKey, Option<Pair<String, String>>> recordLocations = getRecordLocations(hoodieKeyJavaRDD, hoodieTable);
    List<HoodieKey> hoodieKeys = hoodieKeyJavaRDD.collect();
    assertEquals(totalRecords, recordLocations.collect().size());
    assertEquals(totalRecords, recordLocations.map(record -> record._1).distinct().count());
    recordLocations.foreach(entry -> assertTrue(hoodieKeys.contains(entry._1), "Missing HoodieKey"));
    recordLocations.foreach(entry -> assertEquals(recordKeyToPartitionPathMap.get(entry._1.getRecordKey()), entry._1.getPartitionPath(), "PartitionPath mismatch"));
}
Also used : HoodieLayoutConfig(org.apache.hudi.config.HoodieLayoutConfig) HoodieTable(org.apache.hudi.table.HoodieTable) Arrays(java.util.Arrays) HoodieTestDataGenerator(org.apache.hudi.common.testutils.HoodieTestDataGenerator) Random(java.util.Random) HoodieTableType(org.apache.hudi.common.model.HoodieTableType) Assertions.assertFalse(org.junit.jupiter.api.Assertions.assertFalse) Map(java.util.Map) SparkHoodieBackedTableMetadataWriter(org.apache.hudi.metadata.SparkHoodieBackedTableMetadataWriter) SparkBucketIndexPartitioner(org.apache.hudi.table.action.commit.SparkBucketIndexPartitioner) HoodieStorageConfig(org.apache.hudi.config.HoodieStorageConfig) Path(org.apache.hadoop.fs.Path) Tag(org.junit.jupiter.api.Tag) FileSystemViewStorageType(org.apache.hudi.common.table.view.FileSystemViewStorageType) MethodSource(org.junit.jupiter.params.provider.MethodSource) Schema(org.apache.avro.Schema) IndexType(org.apache.hudi.index.HoodieIndex.IndexType) RawTripTestPayload(org.apache.hudi.common.testutils.RawTripTestPayload) UUID(java.util.UUID) Arguments(org.junit.jupiter.params.provider.Arguments) Tuple2(scala.Tuple2) HoodieIndex(org.apache.hudi.index.HoodieIndex) Test(org.junit.jupiter.api.Test) List(java.util.List) Stream(java.util.stream.Stream) FileSystemViewStorageConfig(org.apache.hudi.common.table.view.FileSystemViewStorageConfig) Assertions(org.apache.hudi.testutils.Assertions) Assertions.assertTrue(org.junit.jupiter.api.Assertions.assertTrue) WriteOperationType(org.apache.hudi.common.model.WriteOperationType) MetadataMergeWriteStatus(org.apache.hudi.testutils.MetadataMergeWriteStatus) Assertions.fail(org.junit.jupiter.api.Assertions.fail) Option(org.apache.hudi.common.util.Option) HashMap(java.util.HashMap) ArrayList(java.util.ArrayList) HoodieSparkTable(org.apache.hudi.table.HoodieSparkTable) HoodieTableMetaClient(org.apache.hudi.common.table.HoodieTableMetaClient) Assertions.assertEquals(org.junit.jupiter.api.Assertions.assertEquals) EmptyHoodieRecordPayload(org.apache.hudi.common.model.EmptyHoodieRecordPayload) JavaRDD(org.apache.spark.api.java.JavaRDD) SchemaTestUtil.getSchemaFromResource(org.apache.hudi.common.testutils.SchemaTestUtil.getSchemaFromResource) HoodieMetadataConfig(org.apache.hudi.common.config.HoodieMetadataConfig) ConsistencyGuardConfig(org.apache.hudi.common.fs.ConsistencyGuardConfig) HoodieRecord(org.apache.hudi.common.model.HoodieRecord) Properties(java.util.Properties) HoodieWriteConfig(org.apache.hudi.config.HoodieWriteConfig) IOException(java.io.IOException) JavaPairRDD(org.apache.spark.api.java.JavaPairRDD) HoodieAvroRecord(org.apache.hudi.common.model.HoodieAvroRecord) HoodieCompactionConfig(org.apache.hudi.config.HoodieCompactionConfig) WriteStatus(org.apache.hudi.client.WriteStatus) AfterEach(org.junit.jupiter.api.AfterEach) ParameterizedTest(org.junit.jupiter.params.ParameterizedTest) HoodieIndexConfig(org.apache.hudi.config.HoodieIndexConfig) HoodieKey(org.apache.hudi.common.model.HoodieKey) HoodieTableMetadataWriter(org.apache.hudi.metadata.HoodieTableMetadataWriter) HoodieSparkWriteableTestTable(org.apache.hudi.testutils.HoodieSparkWriteableTestTable) HoodieTestUtils(org.apache.hudi.common.testutils.HoodieTestUtils) Collections(java.util.Collections) Pair(org.apache.hudi.common.util.collection.Pair) HashMap(java.util.HashMap) HoodieRecord(org.apache.hudi.common.model.HoodieRecord) HoodieTable(org.apache.hudi.table.HoodieTable) HoodieKey(org.apache.hudi.common.model.HoodieKey) Option(org.apache.hudi.common.util.Option) MetadataMergeWriteStatus(org.apache.hudi.testutils.MetadataMergeWriteStatus) WriteStatus(org.apache.hudi.client.WriteStatus) ParameterizedTest(org.junit.jupiter.params.ParameterizedTest) MethodSource(org.junit.jupiter.params.provider.MethodSource)

Example 2 with IndexType

use of org.apache.hudi.index.HoodieIndex.IndexType in project hudi by apache.

the class TestHoodieIndex method testTagLocationAndFetchRecordLocations.

@ParameterizedTest
@MethodSource("regularIndexTypeParams")
public void testTagLocationAndFetchRecordLocations(IndexType indexType, boolean populateMetaFields, boolean enableMetadataIndex) throws Exception {
    setUp(indexType, populateMetaFields, enableMetadataIndex);
    String p1 = "2016/01/31";
    String p2 = "2015/01/31";
    String rowKey1 = UUID.randomUUID().toString();
    String rowKey2 = UUID.randomUUID().toString();
    String rowKey3 = UUID.randomUUID().toString();
    String recordStr1 = "{\"_row_key\":\"" + rowKey1 + "\",\"time\":\"2016-01-31T03:16:41.415Z\",\"number\":12}";
    String recordStr2 = "{\"_row_key\":\"" + rowKey2 + "\",\"time\":\"2016-01-31T03:20:41.415Z\",\"number\":100}";
    String recordStr3 = "{\"_row_key\":\"" + rowKey3 + "\",\"time\":\"2016-01-31T03:16:41.415Z\",\"number\":15}";
    // place same row key under a different partition.
    String recordStr4 = "{\"_row_key\":\"" + rowKey1 + "\",\"time\":\"2015-01-31T03:16:41.415Z\",\"number\":32}";
    RawTripTestPayload rowChange1 = new RawTripTestPayload(recordStr1);
    HoodieRecord record1 = new HoodieAvroRecord(new HoodieKey(rowChange1.getRowKey(), rowChange1.getPartitionPath()), rowChange1);
    RawTripTestPayload rowChange2 = new RawTripTestPayload(recordStr2);
    HoodieRecord record2 = new HoodieAvroRecord(new HoodieKey(rowChange2.getRowKey(), rowChange2.getPartitionPath()), rowChange2);
    RawTripTestPayload rowChange3 = new RawTripTestPayload(recordStr3);
    HoodieRecord record3 = new HoodieAvroRecord(new HoodieKey(rowChange3.getRowKey(), rowChange3.getPartitionPath()), rowChange3);
    RawTripTestPayload rowChange4 = new RawTripTestPayload(recordStr4);
    HoodieRecord record4 = new HoodieAvroRecord(new HoodieKey(rowChange4.getRowKey(), rowChange4.getPartitionPath()), rowChange4);
    JavaRDD<HoodieRecord> recordRDD = jsc.parallelize(Arrays.asList(record1, record2, record3, record4));
    String newCommitTime = writeClient.startCommit();
    metaClient = HoodieTableMetaClient.reload(metaClient);
    writeClient.upsert(recordRDD, newCommitTime);
    HoodieTable hoodieTable = HoodieSparkTable.create(config, context, metaClient);
    JavaRDD<HoodieRecord> taggedRecordRDD = tagLocation(index, recordRDD, hoodieTable);
    // Should not find any files
    for (HoodieRecord record : taggedRecordRDD.collect()) {
        assertFalse(record.isCurrentLocationKnown());
    }
    // We create three parquet files, each having one record (two different partitions)
    HoodieSparkWriteableTestTable testTable = HoodieSparkWriteableTestTable.of(metaClient, SCHEMA, metadataWriter);
    final String fileId1 = "fileID1";
    final String fileId2 = "fileID2";
    final String fileId3 = "fileID3";
    Map<String, List<Pair<String, Integer>>> partitionToFilesNameLengthMap = new HashMap<>();
    Path baseFilePath = testTable.forCommit("0000001").withInserts(p1, fileId1, Collections.singletonList(record1));
    long baseFileLength = fs.getFileStatus(baseFilePath).getLen();
    partitionToFilesNameLengthMap.computeIfAbsent(p1, k -> new ArrayList<>()).add(Pair.of(fileId1, Integer.valueOf((int) baseFileLength)));
    testTable.doWriteOperation("0000001", WriteOperationType.UPSERT, Arrays.asList(p1, p2), partitionToFilesNameLengthMap, false, false);
    partitionToFilesNameLengthMap.clear();
    baseFilePath = testTable.forCommit("0000002").withInserts(p1, fileId2, Collections.singletonList(record2));
    baseFileLength = fs.getFileStatus(baseFilePath).getLen();
    partitionToFilesNameLengthMap.computeIfAbsent(p1, k -> new ArrayList<>()).add(Pair.of(fileId2, Integer.valueOf((int) baseFileLength)));
    testTable.doWriteOperation("0000002", WriteOperationType.UPSERT, Arrays.asList(p1, p2), partitionToFilesNameLengthMap, false, false);
    partitionToFilesNameLengthMap.clear();
    baseFilePath = testTable.forCommit("0000003").withInserts(p2, fileId3, Collections.singletonList(record4));
    baseFileLength = fs.getFileStatus(baseFilePath).getLen();
    partitionToFilesNameLengthMap.computeIfAbsent(p2, k -> new ArrayList<>()).add(Pair.of(fileId3, Integer.valueOf((int) baseFileLength)));
    testTable.doWriteOperation("0000003", WriteOperationType.UPSERT, Arrays.asList(p1, p2), partitionToFilesNameLengthMap, false, false);
    // We do the tag again
    metaClient = HoodieTableMetaClient.reload(metaClient);
    hoodieTable = HoodieSparkTable.create(config, context, metaClient);
    taggedRecordRDD = tagLocation(index, recordRDD, hoodieTable);
    List<HoodieRecord> records = taggedRecordRDD.collect();
    // Check results
    for (HoodieRecord record : records) {
        if (record.getRecordKey().equals(rowKey1)) {
            if (record.getPartitionPath().equals(p2)) {
                assertEquals(record.getCurrentLocation().getFileId(), fileId3);
            } else {
                assertEquals(record.getCurrentLocation().getFileId(), fileId1);
            }
        } else if (record.getRecordKey().equals(rowKey2)) {
            assertEquals(record.getCurrentLocation().getFileId(), fileId2);
        } else if (record.getRecordKey().equals(rowKey3)) {
            assertFalse(record.isCurrentLocationKnown());
        }
    }
    JavaPairRDD<HoodieKey, Option<Pair<String, String>>> recordLocations = getRecordLocations(recordRDD.map(HoodieRecord::getKey), hoodieTable);
    for (Tuple2<HoodieKey, Option<Pair<String, String>>> entry : recordLocations.collect()) {
        if (entry._1.getRecordKey().equals(rowKey1)) {
            assertTrue(entry._2.isPresent(), "Row1 should have been present ");
            if (entry._1.getPartitionPath().equals(p2)) {
                assertTrue(entry._2.isPresent(), "Row1 should have been present ");
                assertEquals(entry._2.get().getRight(), fileId3);
            } else {
                assertEquals(entry._2.get().getRight(), fileId1);
            }
        } else if (entry._1.getRecordKey().equals(rowKey2)) {
            assertTrue(entry._2.isPresent(), "Row2 should have been present ");
            assertEquals(entry._2.get().getRight(), fileId2);
        } else if (entry._1.getRecordKey().equals(rowKey3)) {
            assertFalse(entry._2.isPresent(), "Row3 should have been absent ");
        }
    }
}
Also used : Path(org.apache.hadoop.fs.Path) HoodieLayoutConfig(org.apache.hudi.config.HoodieLayoutConfig) HoodieTable(org.apache.hudi.table.HoodieTable) Arrays(java.util.Arrays) HoodieTestDataGenerator(org.apache.hudi.common.testutils.HoodieTestDataGenerator) Random(java.util.Random) HoodieTableType(org.apache.hudi.common.model.HoodieTableType) Assertions.assertFalse(org.junit.jupiter.api.Assertions.assertFalse) Map(java.util.Map) SparkHoodieBackedTableMetadataWriter(org.apache.hudi.metadata.SparkHoodieBackedTableMetadataWriter) SparkBucketIndexPartitioner(org.apache.hudi.table.action.commit.SparkBucketIndexPartitioner) HoodieStorageConfig(org.apache.hudi.config.HoodieStorageConfig) Path(org.apache.hadoop.fs.Path) Tag(org.junit.jupiter.api.Tag) FileSystemViewStorageType(org.apache.hudi.common.table.view.FileSystemViewStorageType) MethodSource(org.junit.jupiter.params.provider.MethodSource) Schema(org.apache.avro.Schema) IndexType(org.apache.hudi.index.HoodieIndex.IndexType) RawTripTestPayload(org.apache.hudi.common.testutils.RawTripTestPayload) UUID(java.util.UUID) Arguments(org.junit.jupiter.params.provider.Arguments) Tuple2(scala.Tuple2) HoodieIndex(org.apache.hudi.index.HoodieIndex) Test(org.junit.jupiter.api.Test) List(java.util.List) Stream(java.util.stream.Stream) FileSystemViewStorageConfig(org.apache.hudi.common.table.view.FileSystemViewStorageConfig) Assertions(org.apache.hudi.testutils.Assertions) Assertions.assertTrue(org.junit.jupiter.api.Assertions.assertTrue) WriteOperationType(org.apache.hudi.common.model.WriteOperationType) MetadataMergeWriteStatus(org.apache.hudi.testutils.MetadataMergeWriteStatus) Assertions.fail(org.junit.jupiter.api.Assertions.fail) Option(org.apache.hudi.common.util.Option) HashMap(java.util.HashMap) ArrayList(java.util.ArrayList) HoodieSparkTable(org.apache.hudi.table.HoodieSparkTable) HoodieTableMetaClient(org.apache.hudi.common.table.HoodieTableMetaClient) Assertions.assertEquals(org.junit.jupiter.api.Assertions.assertEquals) EmptyHoodieRecordPayload(org.apache.hudi.common.model.EmptyHoodieRecordPayload) JavaRDD(org.apache.spark.api.java.JavaRDD) SchemaTestUtil.getSchemaFromResource(org.apache.hudi.common.testutils.SchemaTestUtil.getSchemaFromResource) HoodieMetadataConfig(org.apache.hudi.common.config.HoodieMetadataConfig) ConsistencyGuardConfig(org.apache.hudi.common.fs.ConsistencyGuardConfig) HoodieRecord(org.apache.hudi.common.model.HoodieRecord) Properties(java.util.Properties) HoodieWriteConfig(org.apache.hudi.config.HoodieWriteConfig) IOException(java.io.IOException) JavaPairRDD(org.apache.spark.api.java.JavaPairRDD) HoodieAvroRecord(org.apache.hudi.common.model.HoodieAvroRecord) HoodieCompactionConfig(org.apache.hudi.config.HoodieCompactionConfig) WriteStatus(org.apache.hudi.client.WriteStatus) AfterEach(org.junit.jupiter.api.AfterEach) ParameterizedTest(org.junit.jupiter.params.ParameterizedTest) HoodieIndexConfig(org.apache.hudi.config.HoodieIndexConfig) HoodieKey(org.apache.hudi.common.model.HoodieKey) HoodieTableMetadataWriter(org.apache.hudi.metadata.HoodieTableMetadataWriter) HoodieSparkWriteableTestTable(org.apache.hudi.testutils.HoodieSparkWriteableTestTable) HoodieTestUtils(org.apache.hudi.common.testutils.HoodieTestUtils) Collections(java.util.Collections) Pair(org.apache.hudi.common.util.collection.Pair) HashMap(java.util.HashMap) HoodieRecord(org.apache.hudi.common.model.HoodieRecord) ArrayList(java.util.ArrayList) RawTripTestPayload(org.apache.hudi.common.testutils.RawTripTestPayload) HoodieAvroRecord(org.apache.hudi.common.model.HoodieAvroRecord) HoodieKey(org.apache.hudi.common.model.HoodieKey) HoodieTable(org.apache.hudi.table.HoodieTable) HoodieSparkWriteableTestTable(org.apache.hudi.testutils.HoodieSparkWriteableTestTable) List(java.util.List) ArrayList(java.util.ArrayList) Option(org.apache.hudi.common.util.Option) ParameterizedTest(org.junit.jupiter.params.ParameterizedTest) MethodSource(org.junit.jupiter.params.provider.MethodSource)

Example 3 with IndexType

use of org.apache.hudi.index.HoodieIndex.IndexType in project hudi by apache.

the class TestHoodieClientOnCopyOnWriteStorage method testUpsertsUpdatePartitionPath.

/**
 * This test ensures in a global bloom when update partition path is set to true in config, if an incoming record has mismatched partition
 * compared to whats in storage, then appropriate actions are taken. i.e. old record is deleted in old partition and new one is inserted
 * in the new partition.
 * test structure:
 * 1. insert 1 batch
 * 2. insert 2nd batch with larger no of records so that a new file group is created for partitions
 * 3. issue upserts to records from batch 1 with different partition path. This should ensure records from batch 1 are deleted and new
 * records are upserted to the new partition
 *
 * @param indexType index type to be tested for
 * @param config instance of {@link HoodieWriteConfig} to use
 * @param writeFn write function to be used for testing
 */
private void testUpsertsUpdatePartitionPath(IndexType indexType, HoodieWriteConfig config, Function3<JavaRDD<WriteStatus>, SparkRDDWriteClient, JavaRDD<HoodieRecord>, String> writeFn) throws Exception {
    // instantiate client
    HoodieWriteConfig hoodieWriteConfig = getConfigBuilder().withProps(config.getProps()).withCompactionConfig(HoodieCompactionConfig.newBuilder().compactionSmallFileSize(10000).build()).withIndexConfig(HoodieIndexConfig.newBuilder().withIndexType(indexType).withBloomIndexUpdatePartitionPath(true).withGlobalSimpleIndexUpdatePartitionPath(true).build()).withTimelineLayoutVersion(VERSION_0).build();
    HoodieTableMetaClient.withPropertyBuilder().fromMetaClient(metaClient).setTimelineLayoutVersion(VERSION_0).initTable(metaClient.getHadoopConf(), metaClient.getBasePath());
    // Set rollback to LAZY so no inflights are deleted
    hoodieWriteConfig.getProps().put(HoodieCompactionConfig.FAILED_WRITES_CLEANER_POLICY.key(), HoodieFailedWritesCleaningPolicy.LAZY.name());
    SparkRDDWriteClient client = getHoodieWriteClient(hoodieWriteConfig);
    // Write 1
    String newCommitTime = "001";
    int numRecords = 10;
    client.startCommitWithTime(newCommitTime);
    List<HoodieRecord> records = dataGen.generateInserts(newCommitTime, numRecords);
    Set<Pair<String, String>> expectedPartitionPathRecKeyPairs = new HashSet<>();
    // populate expected partition path and record keys
    for (HoodieRecord rec : records) {
        expectedPartitionPathRecKeyPairs.add(Pair.of(rec.getPartitionPath(), rec.getRecordKey()));
    }
    JavaRDD<HoodieRecord> writeRecords = jsc.parallelize(records, 1);
    JavaRDD<WriteStatus> result = writeFn.apply(client, writeRecords, newCommitTime);
    result.collect();
    // Check the entire dataset has all records
    String[] fullPartitionPaths = getFullPartitionPaths();
    assertPartitionPathRecordKeys(expectedPartitionPathRecKeyPairs, fullPartitionPaths);
    // verify one basefile per partition
    String[] fullExpectedPartitionPaths = getFullPartitionPaths(expectedPartitionPathRecKeyPairs.stream().map(Pair::getLeft).toArray(String[]::new));
    Map<String, Long> baseFileCounts = getBaseFileCountsForPaths(basePath, fs, fullExpectedPartitionPaths);
    for (Map.Entry<String, Long> entry : baseFileCounts.entrySet()) {
        assertEquals(1, entry.getValue());
    }
    assertTrue(baseFileCounts.entrySet().stream().allMatch(entry -> entry.getValue() == 1));
    // Write 2
    newCommitTime = "002";
    // so that a new file id is created
    numRecords = 20;
    client.startCommitWithTime(newCommitTime);
    List<HoodieRecord> recordsSecondBatch = dataGen.generateInserts(newCommitTime, numRecords);
    // populate expected partition path and record keys
    for (HoodieRecord rec : recordsSecondBatch) {
        expectedPartitionPathRecKeyPairs.add(Pair.of(rec.getPartitionPath(), rec.getRecordKey()));
    }
    writeRecords = jsc.parallelize(recordsSecondBatch, 1);
    result = writeFn.apply(client, writeRecords, newCommitTime);
    result.collect();
    // Check the entire dataset has all records
    fullPartitionPaths = getFullPartitionPaths();
    assertPartitionPathRecordKeys(expectedPartitionPathRecKeyPairs, fullPartitionPaths);
    // verify that there are more than 1 basefiles per partition
    // we can't guarantee randomness in partitions where records are distributed. So, verify atleast one partition has more than 1 basefile.
    baseFileCounts = getBaseFileCountsForPaths(basePath, fs, fullPartitionPaths);
    assertTrue(baseFileCounts.entrySet().stream().filter(entry -> entry.getValue() > 1).count() >= 1, "At least one partition should have more than 1 base file after 2nd batch of writes");
    // Write 3 (upserts to records from batch 1 with diff partition path)
    newCommitTime = "003";
    // update to diff partition paths
    List<HoodieRecord> recordsToUpsert = new ArrayList<>();
    for (HoodieRecord rec : records) {
        // remove older entry from expected partition path record key pairs
        expectedPartitionPathRecKeyPairs.remove(Pair.of(rec.getPartitionPath(), rec.getRecordKey()));
        String partitionPath = rec.getPartitionPath();
        String newPartitionPath = null;
        if (partitionPath.equalsIgnoreCase(DEFAULT_FIRST_PARTITION_PATH)) {
            newPartitionPath = DEFAULT_SECOND_PARTITION_PATH;
        } else if (partitionPath.equalsIgnoreCase(DEFAULT_SECOND_PARTITION_PATH)) {
            newPartitionPath = DEFAULT_THIRD_PARTITION_PATH;
        } else if (partitionPath.equalsIgnoreCase(DEFAULT_THIRD_PARTITION_PATH)) {
            newPartitionPath = DEFAULT_FIRST_PARTITION_PATH;
        } else {
            throw new IllegalStateException("Unknown partition path " + rec.getPartitionPath());
        }
        recordsToUpsert.add(new HoodieAvroRecord(new HoodieKey(rec.getRecordKey(), newPartitionPath), (HoodieRecordPayload) rec.getData()));
        // populate expected partition path and record keys
        expectedPartitionPathRecKeyPairs.add(Pair.of(newPartitionPath, rec.getRecordKey()));
    }
    writeRecords = jsc.parallelize(recordsToUpsert, 1);
    result = writeFn.apply(client, writeRecords, newCommitTime);
    result.collect();
    // Check the entire dataset has all records
    fullPartitionPaths = getFullPartitionPaths();
    assertPartitionPathRecordKeys(expectedPartitionPathRecKeyPairs, fullPartitionPaths);
}
Also used : HoodieTable(org.apache.hudi.table.HoodieTable) BeforeEach(org.junit.jupiter.api.BeforeEach) Arrays(java.util.Arrays) FileIOUtils(org.apache.hudi.common.util.FileIOUtils) HoodieUpsertException(org.apache.hudi.exception.HoodieUpsertException) SparkSingleFileSortPlanStrategy(org.apache.hudi.client.clustering.plan.strategy.SparkSingleFileSortPlanStrategy) SparkTaskContextSupplier(org.apache.hudi.client.SparkTaskContextSupplier) HoodieWriteHelper(org.apache.hudi.table.action.commit.HoodieWriteHelper) BaseKeyGenerator(org.apache.hudi.keygen.BaseKeyGenerator) Future(java.util.concurrent.Future) Map(java.util.Map) EAGER(org.apache.hudi.common.model.HoodieFailedWritesCleaningPolicy.EAGER) Tag(org.junit.jupiter.api.Tag) HoodieWriteResult(org.apache.hudi.client.HoodieWriteResult) REQUESTED(org.apache.hudi.common.table.timeline.HoodieInstant.State.REQUESTED) HoodieWriteMetadata(org.apache.hudi.table.action.HoodieWriteMetadata) HoodieFileGroupId(org.apache.hudi.common.model.HoodieFileGroupId) HoodieActiveTimeline(org.apache.hudi.common.table.timeline.HoodieActiveTimeline) FSDataInputStream(org.apache.hadoop.fs.FSDataInputStream) BaseHoodieWriteClient(org.apache.hudi.client.BaseHoodieWriteClient) IndexType(org.apache.hudi.index.HoodieIndex.IndexType) HoodieClusteringPlan(org.apache.hudi.avro.model.HoodieClusteringPlan) Set(java.util.Set) VERSION_0(org.apache.hudi.common.table.timeline.versioning.TimelineLayoutVersion.VERSION_0) Arguments(org.junit.jupiter.params.provider.Arguments) HoodieIndex(org.apache.hudi.index.HoodieIndex) Executors(java.util.concurrent.Executors) HoodieBaseFile(org.apache.hudi.common.model.HoodieBaseFile) Stream(java.util.stream.Stream) FileSystemViewStorageConfig(org.apache.hudi.common.table.view.FileSystemViewStorageConfig) Assertions.assertTrue(org.junit.jupiter.api.Assertions.assertTrue) ClusteringUtils(org.apache.hudi.common.util.ClusteringUtils) HoodieClientTestUtils(org.apache.hudi.testutils.HoodieClientTestUtils) SqlQuerySingleResultPreCommitValidator(org.apache.hudi.client.validator.SqlQuerySingleResultPreCommitValidator) DEFAULT_THIRD_PARTITION_PATH(org.apache.hudi.common.testutils.HoodieTestDataGenerator.DEFAULT_THIRD_PARTITION_PATH) Mockito.mock(org.mockito.Mockito.mock) HoodieClientTestBase(org.apache.hudi.testutils.HoodieClientTestBase) Assertions.assertThrows(org.junit.jupiter.api.Assertions.assertThrows) Assertions.fail(org.junit.jupiter.api.Assertions.fail) Dataset(org.apache.spark.sql.Dataset) Assertions.assertNull(org.junit.jupiter.api.Assertions.assertNull) Option(org.apache.hudi.common.util.Option) HoodieEngineContext(org.apache.hudi.common.engine.HoodieEngineContext) DEFAULT_FIRST_PARTITION_PATH(org.apache.hudi.common.testutils.HoodieTestDataGenerator.DEFAULT_FIRST_PARTITION_PATH) HoodieValidationException(org.apache.hudi.exception.HoodieValidationException) ArrayList(java.util.ArrayList) MarkerType(org.apache.hudi.common.table.marker.MarkerType) StringUtils(org.apache.hudi.common.util.StringUtils) KeyGenerator(org.apache.hudi.keygen.KeyGenerator) BulkInsertPartitioner(org.apache.hudi.table.BulkInsertPartitioner) Transformations.recordsToRecordKeySet(org.apache.hudi.common.testutils.Transformations.recordsToRecordKeySet) EXECUTION_STRATEGY_CLASS_NAME(org.apache.hudi.config.HoodieClusteringConfig.EXECUTION_STRATEGY_CLASS_NAME) Assertions.assertEquals(org.junit.jupiter.api.Assertions.assertEquals) JavaRDD(org.apache.spark.api.java.JavaRDD) TimelineLayoutVersion(org.apache.hudi.common.table.timeline.versioning.TimelineLayoutVersion) ValueSource(org.junit.jupiter.params.provider.ValueSource) ConsistencyGuardConfig(org.apache.hudi.common.fs.ConsistencyGuardConfig) HoodieRecord(org.apache.hudi.common.model.HoodieRecord) Assertions.assertNoWriteErrors(org.apache.hudi.testutils.Assertions.assertNoWriteErrors) HoodieData(org.apache.hudi.common.data.HoodieData) RDDCustomColumnsSortPartitioner(org.apache.hudi.execution.bulkinsert.RDDCustomColumnsSortPartitioner) Properties(java.util.Properties) HoodieWriteConfig(org.apache.hudi.config.HoodieWriteConfig) WriteMarkersFactory(org.apache.hudi.table.marker.WriteMarkersFactory) BaseFileOnlyView(org.apache.hudi.common.table.view.TableFileSystemView.BaseFileOnlyView) SqlQueryEqualityPreCommitValidator(org.apache.hudi.client.validator.SqlQueryEqualityPreCommitValidator) DEFAULT_SECOND_PARTITION_PATH(org.apache.hudi.common.testutils.HoodieTestDataGenerator.DEFAULT_SECOND_PARTITION_PATH) HoodieTestTable(org.apache.hudi.common.testutils.HoodieTestTable) HoodieCommitMetadata(org.apache.hudi.common.model.HoodieCommitMetadata) IOException(java.io.IOException) Row(org.apache.spark.sql.Row) HoodieAvroRecord(org.apache.hudi.common.model.HoodieAvroRecord) HoodieCompactionConfig(org.apache.hudi.config.HoodieCompactionConfig) ParameterizedTest(org.junit.jupiter.params.ParameterizedTest) HoodieCleanMetadata(org.apache.hudi.avro.model.HoodieCleanMetadata) HoodieCorruptedDataException(org.apache.hudi.exception.HoodieCorruptedDataException) HoodieKey(org.apache.hudi.common.model.HoodieKey) HoodieSparkWriteableTestTable(org.apache.hudi.testutils.HoodieSparkWriteableTestTable) HoodieIOException(org.apache.hudi.exception.HoodieIOException) HoodieTestUtils(org.apache.hudi.common.testutils.HoodieTestUtils) COMPLETED(org.apache.hudi.common.table.timeline.HoodieInstant.State.COMPLETED) REPLACE_COMMIT_ACTION(org.apache.hudi.common.table.timeline.HoodieTimeline.REPLACE_COMMIT_ACTION) HoodieFailedWritesCleaningPolicy(org.apache.hudi.common.model.HoodieFailedWritesCleaningPolicy) HoodieInstant(org.apache.hudi.common.table.timeline.HoodieInstant) HoodieTestDataGenerator(org.apache.hudi.common.testutils.HoodieTestDataGenerator) CollectionUtils(org.apache.hudi.common.util.CollectionUtils) HoodieJavaRDD(org.apache.hudi.data.HoodieJavaRDD) Logger(org.apache.log4j.Logger) HoodieMergeHandle(org.apache.hudi.io.HoodieMergeHandle) CLEAN_ACTION(org.apache.hudi.common.table.timeline.HoodieTimeline.CLEAN_ACTION) Assertions.assertFalse(org.junit.jupiter.api.Assertions.assertFalse) HoodieStorageConfig(org.apache.hudi.config.HoodieStorageConfig) Path(org.apache.hadoop.fs.Path) HoodieSparkKeyGeneratorFactory(org.apache.hudi.keygen.factory.HoodieSparkKeyGeneratorFactory) MethodSource(org.junit.jupiter.params.provider.MethodSource) HoodieRollbackException(org.apache.hudi.exception.HoodieRollbackException) SparkSingleFileSortExecutionStrategy(org.apache.hudi.client.clustering.run.strategy.SparkSingleFileSortExecutionStrategy) HoodiePreCommitValidatorConfig(org.apache.hudi.config.HoodiePreCommitValidatorConfig) TRIP_EXAMPLE_SCHEMA(org.apache.hudi.common.testutils.HoodieTestDataGenerator.TRIP_EXAMPLE_SCHEMA) IOType(org.apache.hudi.common.model.IOType) RawTripTestPayload(org.apache.hudi.common.testutils.RawTripTestPayload) Collection(java.util.Collection) TimelineMetadataUtils(org.apache.hudi.common.table.timeline.TimelineMetadataUtils) MarkerUtils(org.apache.hudi.common.util.MarkerUtils) UUID(java.util.UUID) Collectors(java.util.stream.Collectors) Test(org.junit.jupiter.api.Test) List(java.util.List) FileCreateUtils.getBaseFileCountsForPaths(org.apache.hudi.common.testutils.FileCreateUtils.getBaseFileCountsForPaths) HoodieWriteStat(org.apache.hudi.common.model.HoodieWriteStat) ROLLBACK_ACTION(org.apache.hudi.common.table.timeline.HoodieTimeline.ROLLBACK_ACTION) WriteOperationType(org.apache.hudi.common.model.WriteOperationType) NotNull(org.jetbrains.annotations.NotNull) HoodieInsertException(org.apache.hudi.exception.HoodieInsertException) Transformations.randomSelectAsHoodieKeys(org.apache.hudi.common.testutils.Transformations.randomSelectAsHoodieKeys) INFLIGHT(org.apache.hudi.common.table.timeline.HoodieInstant.State.INFLIGHT) COMMIT_ACTION(org.apache.hudi.common.table.timeline.HoodieTimeline.COMMIT_ACTION) BaseFileUtils(org.apache.hudi.common.util.BaseFileUtils) FileSlice(org.apache.hudi.common.model.FileSlice) HoodieCommitException(org.apache.hudi.exception.HoodieCommitException) EnumSource(org.junit.jupiter.params.provider.EnumSource) HashMap(java.util.HashMap) HashSet(java.util.HashSet) HoodieSparkTable(org.apache.hudi.table.HoodieSparkTable) HoodieRequestedReplaceMetadata(org.apache.hudi.avro.model.HoodieRequestedReplaceMetadata) HoodieTableMetaClient(org.apache.hudi.common.table.HoodieTableMetaClient) HoodieSparkCopyOnWriteTable(org.apache.hudi.table.HoodieSparkCopyOnWriteTable) HoodieTimeline(org.apache.hudi.common.table.timeline.HoodieTimeline) ExecutorService(java.util.concurrent.ExecutorService) GenericRecord(org.apache.avro.generic.GenericRecord) ASYNC_CLUSTERING_ENABLE(org.apache.hudi.config.HoodieClusteringConfig.ASYNC_CLUSTERING_ENABLE) TypedProperties(org.apache.hudi.common.config.TypedProperties) NULL_SCHEMA(org.apache.hudi.common.testutils.HoodieTestDataGenerator.NULL_SCHEMA) Mockito.when(org.mockito.Mockito.when) FileCreateUtils(org.apache.hudi.common.testutils.FileCreateUtils) WriteStatus(org.apache.hudi.client.WriteStatus) HoodieRecordPayload(org.apache.hudi.common.model.HoodieRecordPayload) ClusteringTestUtils(org.apache.hudi.common.testutils.ClusteringTestUtils) SparkPreCommitValidator(org.apache.hudi.client.validator.SparkPreCommitValidator) SparkRDDWriteClient(org.apache.hudi.client.SparkRDDWriteClient) HoodieIndexConfig(org.apache.hudi.config.HoodieIndexConfig) HoodieClusteringConfig(org.apache.hudi.config.HoodieClusteringConfig) LogManager(org.apache.log4j.LogManager) Collections(java.util.Collections) FSUtils(org.apache.hudi.common.fs.FSUtils) Pair(org.apache.hudi.common.util.collection.Pair) SparkRDDWriteClient(org.apache.hudi.client.SparkRDDWriteClient) HoodieRecord(org.apache.hudi.common.model.HoodieRecord) ArrayList(java.util.ArrayList) HoodieWriteConfig(org.apache.hudi.config.HoodieWriteConfig) HoodieRecordPayload(org.apache.hudi.common.model.HoodieRecordPayload) HoodieAvroRecord(org.apache.hudi.common.model.HoodieAvroRecord) HoodieKey(org.apache.hudi.common.model.HoodieKey) Map(java.util.Map) HashMap(java.util.HashMap) WriteStatus(org.apache.hudi.client.WriteStatus) Pair(org.apache.hudi.common.util.collection.Pair) HashSet(java.util.HashSet)

Example 4 with IndexType

use of org.apache.hudi.index.HoodieIndex.IndexType in project hudi by apache.

the class TestHoodieIndex method testTagLocationAndDuplicateUpdate.

@ParameterizedTest
@MethodSource("indexTypeParams")
public void testTagLocationAndDuplicateUpdate(IndexType indexType, boolean populateMetaFields, boolean enableMetadataIndex) throws Exception {
    setUp(indexType, populateMetaFields, enableMetadataIndex);
    String newCommitTime = "001";
    int totalRecords = 10 + random.nextInt(20);
    List<HoodieRecord> records = dataGen.generateInserts(newCommitTime, totalRecords);
    JavaRDD<HoodieRecord> writeRecords = jsc.parallelize(records, 1);
    HoodieSparkTable hoodieTable = HoodieSparkTable.create(config, context, metaClient);
    writeClient.startCommitWithTime(newCommitTime);
    JavaRDD<WriteStatus> writeStatues = writeClient.upsert(writeRecords, newCommitTime);
    JavaRDD<HoodieRecord> javaRDD1 = tagLocation(index, writeRecords, hoodieTable);
    // Duplicate upsert and ensure correctness is maintained
    // We are trying to approximately imitate the case when the RDD is recomputed. For RDD creating, driver code is not
    // recomputed. This includes the state transitions. We need to delete the inflight instance so that subsequent
    // upsert will not run into conflicts.
    metaClient.getFs().delete(new Path(metaClient.getMetaPath(), "001.inflight"));
    writeClient.upsert(writeRecords, newCommitTime);
    Assertions.assertNoWriteErrors(writeStatues.collect());
    // Now commit this & update location of records inserted and validate no errors
    writeClient.commit(newCommitTime, writeStatues);
    // Now tagLocation for these records, hbaseIndex should tag them correctly
    metaClient = HoodieTableMetaClient.reload(metaClient);
    hoodieTable = HoodieSparkTable.create(config, context, metaClient);
    JavaRDD<HoodieRecord> javaRDD = tagLocation(index, writeRecords, hoodieTable);
    Map<String, String> recordKeyToPartitionPathMap = new HashMap();
    List<HoodieRecord> hoodieRecords = writeRecords.collect();
    hoodieRecords.forEach(entry -> recordKeyToPartitionPathMap.put(entry.getRecordKey(), entry.getPartitionPath()));
    assertEquals(totalRecords, javaRDD.filter(HoodieRecord::isCurrentLocationKnown).collect().size());
    assertEquals(totalRecords, javaRDD.map(record -> record.getKey().getRecordKey()).distinct().count());
    assertEquals(totalRecords, javaRDD.filter(record -> (record.getCurrentLocation() != null && record.getCurrentLocation().getInstantTime().equals(newCommitTime))).distinct().count());
    javaRDD.foreach(entry -> assertEquals(recordKeyToPartitionPathMap.get(entry.getRecordKey()), entry.getPartitionPath(), "PartitionPath mismatch"));
    JavaRDD<HoodieKey> hoodieKeyJavaRDD = writeRecords.map(entry -> entry.getKey());
    JavaPairRDD<HoodieKey, Option<Pair<String, String>>> recordLocations = getRecordLocations(hoodieKeyJavaRDD, hoodieTable);
    List<HoodieKey> hoodieKeys = hoodieKeyJavaRDD.collect();
    assertEquals(totalRecords, recordLocations.collect().size());
    assertEquals(totalRecords, recordLocations.map(record -> record._1).distinct().count());
    recordLocations.foreach(entry -> assertTrue(hoodieKeys.contains(entry._1), "Missing HoodieKey"));
    recordLocations.foreach(entry -> assertEquals(recordKeyToPartitionPathMap.get(entry._1.getRecordKey()), entry._1.getPartitionPath(), "PartitionPath mismatch"));
}
Also used : Path(org.apache.hadoop.fs.Path) HoodieLayoutConfig(org.apache.hudi.config.HoodieLayoutConfig) HoodieTable(org.apache.hudi.table.HoodieTable) Arrays(java.util.Arrays) HoodieTestDataGenerator(org.apache.hudi.common.testutils.HoodieTestDataGenerator) Random(java.util.Random) HoodieTableType(org.apache.hudi.common.model.HoodieTableType) Assertions.assertFalse(org.junit.jupiter.api.Assertions.assertFalse) Map(java.util.Map) SparkHoodieBackedTableMetadataWriter(org.apache.hudi.metadata.SparkHoodieBackedTableMetadataWriter) SparkBucketIndexPartitioner(org.apache.hudi.table.action.commit.SparkBucketIndexPartitioner) HoodieStorageConfig(org.apache.hudi.config.HoodieStorageConfig) Path(org.apache.hadoop.fs.Path) Tag(org.junit.jupiter.api.Tag) FileSystemViewStorageType(org.apache.hudi.common.table.view.FileSystemViewStorageType) MethodSource(org.junit.jupiter.params.provider.MethodSource) Schema(org.apache.avro.Schema) IndexType(org.apache.hudi.index.HoodieIndex.IndexType) RawTripTestPayload(org.apache.hudi.common.testutils.RawTripTestPayload) UUID(java.util.UUID) Arguments(org.junit.jupiter.params.provider.Arguments) Tuple2(scala.Tuple2) HoodieIndex(org.apache.hudi.index.HoodieIndex) Test(org.junit.jupiter.api.Test) List(java.util.List) Stream(java.util.stream.Stream) FileSystemViewStorageConfig(org.apache.hudi.common.table.view.FileSystemViewStorageConfig) Assertions(org.apache.hudi.testutils.Assertions) Assertions.assertTrue(org.junit.jupiter.api.Assertions.assertTrue) WriteOperationType(org.apache.hudi.common.model.WriteOperationType) MetadataMergeWriteStatus(org.apache.hudi.testutils.MetadataMergeWriteStatus) Assertions.fail(org.junit.jupiter.api.Assertions.fail) Option(org.apache.hudi.common.util.Option) HashMap(java.util.HashMap) ArrayList(java.util.ArrayList) HoodieSparkTable(org.apache.hudi.table.HoodieSparkTable) HoodieTableMetaClient(org.apache.hudi.common.table.HoodieTableMetaClient) Assertions.assertEquals(org.junit.jupiter.api.Assertions.assertEquals) EmptyHoodieRecordPayload(org.apache.hudi.common.model.EmptyHoodieRecordPayload) JavaRDD(org.apache.spark.api.java.JavaRDD) SchemaTestUtil.getSchemaFromResource(org.apache.hudi.common.testutils.SchemaTestUtil.getSchemaFromResource) HoodieMetadataConfig(org.apache.hudi.common.config.HoodieMetadataConfig) ConsistencyGuardConfig(org.apache.hudi.common.fs.ConsistencyGuardConfig) HoodieRecord(org.apache.hudi.common.model.HoodieRecord) Properties(java.util.Properties) HoodieWriteConfig(org.apache.hudi.config.HoodieWriteConfig) IOException(java.io.IOException) JavaPairRDD(org.apache.spark.api.java.JavaPairRDD) HoodieAvroRecord(org.apache.hudi.common.model.HoodieAvroRecord) HoodieCompactionConfig(org.apache.hudi.config.HoodieCompactionConfig) WriteStatus(org.apache.hudi.client.WriteStatus) AfterEach(org.junit.jupiter.api.AfterEach) ParameterizedTest(org.junit.jupiter.params.ParameterizedTest) HoodieIndexConfig(org.apache.hudi.config.HoodieIndexConfig) HoodieKey(org.apache.hudi.common.model.HoodieKey) HoodieTableMetadataWriter(org.apache.hudi.metadata.HoodieTableMetadataWriter) HoodieSparkWriteableTestTable(org.apache.hudi.testutils.HoodieSparkWriteableTestTable) HoodieTestUtils(org.apache.hudi.common.testutils.HoodieTestUtils) Collections(java.util.Collections) Pair(org.apache.hudi.common.util.collection.Pair) HashMap(java.util.HashMap) HoodieRecord(org.apache.hudi.common.model.HoodieRecord) HoodieKey(org.apache.hudi.common.model.HoodieKey) Option(org.apache.hudi.common.util.Option) MetadataMergeWriteStatus(org.apache.hudi.testutils.MetadataMergeWriteStatus) WriteStatus(org.apache.hudi.client.WriteStatus) HoodieSparkTable(org.apache.hudi.table.HoodieSparkTable) ParameterizedTest(org.junit.jupiter.params.ParameterizedTest) MethodSource(org.junit.jupiter.params.provider.MethodSource)

Example 5 with IndexType

use of org.apache.hudi.index.HoodieIndex.IndexType in project hudi by apache.

the class TestHoodieIndex method testSimpleTagLocationAndUpdateWithRollback.

@ParameterizedTest
@MethodSource("indexTypeParams")
public void testSimpleTagLocationAndUpdateWithRollback(IndexType indexType, boolean populateMetaFields, boolean enableMetadataIndex) throws Exception {
    setUp(indexType, populateMetaFields, false, enableMetadataIndex);
    String newCommitTime = writeClient.startCommit();
    int totalRecords = 20 + random.nextInt(20);
    List<HoodieRecord> records = dataGen.generateInserts(newCommitTime, totalRecords);
    JavaRDD<HoodieRecord> writeRecords = jsc.parallelize(records, 1);
    metaClient = HoodieTableMetaClient.reload(metaClient);
    // Insert 200 records
    JavaRDD<WriteStatus> writeStatues = writeClient.upsert(writeRecords, newCommitTime);
    Assertions.assertNoWriteErrors(writeStatues.collect());
    List<String> fileIds = writeStatues.map(WriteStatus::getFileId).collect();
    // commit this upsert
    writeClient.commit(newCommitTime, writeStatues);
    HoodieTable hoodieTable = HoodieSparkTable.create(config, context, metaClient);
    // Now tagLocation for these records, hbaseIndex should tag them
    JavaRDD<HoodieRecord> javaRDD = tagLocation(index, writeRecords, hoodieTable);
    assert (javaRDD.filter(HoodieRecord::isCurrentLocationKnown).collect().size() == totalRecords);
    // check tagged records are tagged with correct fileIds
    assert (javaRDD.filter(record -> record.getCurrentLocation().getFileId() == null).collect().size() == 0);
    List<String> taggedFileIds = javaRDD.map(record -> record.getCurrentLocation().getFileId()).distinct().collect();
    Map<String, String> recordKeyToPartitionPathMap = new HashMap();
    List<HoodieRecord> hoodieRecords = writeRecords.collect();
    hoodieRecords.forEach(entry -> recordKeyToPartitionPathMap.put(entry.getRecordKey(), entry.getPartitionPath()));
    JavaRDD<HoodieKey> hoodieKeyJavaRDD = writeRecords.map(entry -> entry.getKey());
    JavaPairRDD<HoodieKey, Option<Pair<String, String>>> recordLocations = getRecordLocations(hoodieKeyJavaRDD, hoodieTable);
    List<HoodieKey> hoodieKeys = hoodieKeyJavaRDD.collect();
    assertEquals(totalRecords, recordLocations.collect().size());
    assertEquals(totalRecords, recordLocations.map(record -> record._1).distinct().count());
    recordLocations.foreach(entry -> assertTrue(hoodieKeys.contains(entry._1), "Missing HoodieKey"));
    recordLocations.foreach(entry -> assertEquals(recordKeyToPartitionPathMap.get(entry._1.getRecordKey()), entry._1.getPartitionPath(), "PartitionPath mismatch"));
    // both lists should match
    assertTrue(taggedFileIds.containsAll(fileIds) && fileIds.containsAll(taggedFileIds));
    // Rollback the last commit
    writeClient.rollback(newCommitTime);
    hoodieTable = HoodieSparkTable.create(config, context, metaClient);
    // Now tagLocation for these records, hbaseIndex should not tag them since it was a rolled
    // back commit
    javaRDD = tagLocation(index, writeRecords, hoodieTable);
    assert (javaRDD.filter(HoodieRecord::isCurrentLocationKnown).collect().size() == 0);
    assert (javaRDD.filter(record -> record.getCurrentLocation() != null).collect().size() == 0);
}
Also used : HoodieLayoutConfig(org.apache.hudi.config.HoodieLayoutConfig) HoodieTable(org.apache.hudi.table.HoodieTable) Arrays(java.util.Arrays) HoodieTestDataGenerator(org.apache.hudi.common.testutils.HoodieTestDataGenerator) Random(java.util.Random) HoodieTableType(org.apache.hudi.common.model.HoodieTableType) Assertions.assertFalse(org.junit.jupiter.api.Assertions.assertFalse) Map(java.util.Map) SparkHoodieBackedTableMetadataWriter(org.apache.hudi.metadata.SparkHoodieBackedTableMetadataWriter) SparkBucketIndexPartitioner(org.apache.hudi.table.action.commit.SparkBucketIndexPartitioner) HoodieStorageConfig(org.apache.hudi.config.HoodieStorageConfig) Path(org.apache.hadoop.fs.Path) Tag(org.junit.jupiter.api.Tag) FileSystemViewStorageType(org.apache.hudi.common.table.view.FileSystemViewStorageType) MethodSource(org.junit.jupiter.params.provider.MethodSource) Schema(org.apache.avro.Schema) IndexType(org.apache.hudi.index.HoodieIndex.IndexType) RawTripTestPayload(org.apache.hudi.common.testutils.RawTripTestPayload) UUID(java.util.UUID) Arguments(org.junit.jupiter.params.provider.Arguments) Tuple2(scala.Tuple2) HoodieIndex(org.apache.hudi.index.HoodieIndex) Test(org.junit.jupiter.api.Test) List(java.util.List) Stream(java.util.stream.Stream) FileSystemViewStorageConfig(org.apache.hudi.common.table.view.FileSystemViewStorageConfig) Assertions(org.apache.hudi.testutils.Assertions) Assertions.assertTrue(org.junit.jupiter.api.Assertions.assertTrue) WriteOperationType(org.apache.hudi.common.model.WriteOperationType) MetadataMergeWriteStatus(org.apache.hudi.testutils.MetadataMergeWriteStatus) Assertions.fail(org.junit.jupiter.api.Assertions.fail) Option(org.apache.hudi.common.util.Option) HashMap(java.util.HashMap) ArrayList(java.util.ArrayList) HoodieSparkTable(org.apache.hudi.table.HoodieSparkTable) HoodieTableMetaClient(org.apache.hudi.common.table.HoodieTableMetaClient) Assertions.assertEquals(org.junit.jupiter.api.Assertions.assertEquals) EmptyHoodieRecordPayload(org.apache.hudi.common.model.EmptyHoodieRecordPayload) JavaRDD(org.apache.spark.api.java.JavaRDD) SchemaTestUtil.getSchemaFromResource(org.apache.hudi.common.testutils.SchemaTestUtil.getSchemaFromResource) HoodieMetadataConfig(org.apache.hudi.common.config.HoodieMetadataConfig) ConsistencyGuardConfig(org.apache.hudi.common.fs.ConsistencyGuardConfig) HoodieRecord(org.apache.hudi.common.model.HoodieRecord) Properties(java.util.Properties) HoodieWriteConfig(org.apache.hudi.config.HoodieWriteConfig) IOException(java.io.IOException) JavaPairRDD(org.apache.spark.api.java.JavaPairRDD) HoodieAvroRecord(org.apache.hudi.common.model.HoodieAvroRecord) HoodieCompactionConfig(org.apache.hudi.config.HoodieCompactionConfig) WriteStatus(org.apache.hudi.client.WriteStatus) AfterEach(org.junit.jupiter.api.AfterEach) ParameterizedTest(org.junit.jupiter.params.ParameterizedTest) HoodieIndexConfig(org.apache.hudi.config.HoodieIndexConfig) HoodieKey(org.apache.hudi.common.model.HoodieKey) HoodieTableMetadataWriter(org.apache.hudi.metadata.HoodieTableMetadataWriter) HoodieSparkWriteableTestTable(org.apache.hudi.testutils.HoodieSparkWriteableTestTable) HoodieTestUtils(org.apache.hudi.common.testutils.HoodieTestUtils) Collections(java.util.Collections) Pair(org.apache.hudi.common.util.collection.Pair) HashMap(java.util.HashMap) HoodieRecord(org.apache.hudi.common.model.HoodieRecord) HoodieTable(org.apache.hudi.table.HoodieTable) HoodieKey(org.apache.hudi.common.model.HoodieKey) Option(org.apache.hudi.common.util.Option) MetadataMergeWriteStatus(org.apache.hudi.testutils.MetadataMergeWriteStatus) WriteStatus(org.apache.hudi.client.WriteStatus) ParameterizedTest(org.junit.jupiter.params.ParameterizedTest) MethodSource(org.junit.jupiter.params.provider.MethodSource)

Aggregations

IOException (java.io.IOException)5 ArrayList (java.util.ArrayList)5 Arrays (java.util.Arrays)5 Collections (java.util.Collections)5 HashMap (java.util.HashMap)5 List (java.util.List)5 Map (java.util.Map)5 Properties (java.util.Properties)5 UUID (java.util.UUID)5 Stream (java.util.stream.Stream)5 Path (org.apache.hadoop.fs.Path)5 WriteStatus (org.apache.hudi.client.WriteStatus)5 ConsistencyGuardConfig (org.apache.hudi.common.fs.ConsistencyGuardConfig)5 HoodieAvroRecord (org.apache.hudi.common.model.HoodieAvroRecord)5 HoodieKey (org.apache.hudi.common.model.HoodieKey)5 HoodieRecord (org.apache.hudi.common.model.HoodieRecord)5 WriteOperationType (org.apache.hudi.common.model.WriteOperationType)5 HoodieTableMetaClient (org.apache.hudi.common.table.HoodieTableMetaClient)5 FileSystemViewStorageConfig (org.apache.hudi.common.table.view.FileSystemViewStorageConfig)5 HoodieTestDataGenerator (org.apache.hudi.common.testutils.HoodieTestDataGenerator)5