Examples with HoodieDeleteBlock org.apache.hudi.common.table.log.block.HoodieDeleteBlock used on opensource projects

Example 6 with HoodieDeleteBlock

use of org.apache.hudi.common.table.log.block.HoodieDeleteBlock in project hudi by apache.

the class TestHoodieLogFormat method testAvroLogRecordReaderWithDeleteAndRollback.

@ParameterizedTest
@MethodSource("testArguments")
public void testAvroLogRecordReaderWithDeleteAndRollback(ExternalSpillableMap.DiskMapType diskMapType, boolean isCompressionEnabled, boolean readBlocksLazily) throws IOException, URISyntaxException, InterruptedException {
    Schema schema = HoodieAvroUtils.addMetadataFields(getSimpleSchema());
    // Set a small threshold so that every block is a new version
    Writer writer = HoodieLogFormat.newWriterBuilder().onParentPath(partitionPath).withFileExtension(HoodieLogFile.DELTA_EXTENSION).withFileId("test-fileid1").overBaseCommit("100").withFs(fs).build();
    // Write 1
    List<IndexedRecord> records1 = SchemaTestUtil.generateHoodieTestRecords(0, 100);
    List<IndexedRecord> copyOfRecords1 = records1.stream().map(record -> HoodieAvroUtils.rewriteRecord((GenericRecord) record, schema)).collect(Collectors.toList());
    Map<HoodieLogBlock.HeaderMetadataType, String> header = new HashMap<>();
    header.put(HoodieLogBlock.HeaderMetadataType.INSTANT_TIME, "100");
    header.put(HoodieLogBlock.HeaderMetadataType.SCHEMA, schema.toString());
    HoodieDataBlock dataBlock = getDataBlock(DEFAULT_DATA_BLOCK_TYPE, records1, header);
    writer.appendBlock(dataBlock);
    // Write 2
    header.put(HoodieLogBlock.HeaderMetadataType.INSTANT_TIME, "101");
    List<IndexedRecord> records2 = SchemaTestUtil.generateHoodieTestRecords(0, 100);
    List<IndexedRecord> copyOfRecords2 = records2.stream().map(record -> HoodieAvroUtils.rewriteRecord((GenericRecord) record, schema)).collect(Collectors.toList());
    dataBlock = getDataBlock(DEFAULT_DATA_BLOCK_TYPE, records2, header);
    writer.appendBlock(dataBlock);
    copyOfRecords1.addAll(copyOfRecords2);
    List<String> originalKeys = copyOfRecords1.stream().map(s -> ((GenericRecord) s).get(HoodieRecord.RECORD_KEY_METADATA_FIELD).toString()).collect(Collectors.toList());
    // Delete 50 keys
    List<HoodieKey> deletedKeys = copyOfRecords1.stream().map(s -> (new HoodieKey(((GenericRecord) s).get(HoodieRecord.RECORD_KEY_METADATA_FIELD).toString(), ((GenericRecord) s).get(HoodieRecord.PARTITION_PATH_METADATA_FIELD).toString()))).collect(Collectors.toList()).subList(0, 50);
    header.put(HoodieLogBlock.HeaderMetadataType.INSTANT_TIME, "102");
    HoodieDeleteBlock deleteBlock = new HoodieDeleteBlock(deletedKeys.toArray(new HoodieKey[50]), header);
    writer.appendBlock(deleteBlock);
    List<String> allLogFiles = FSUtils.getAllLogFiles(fs, partitionPath, "test-fileid1", HoodieLogFile.DELTA_EXTENSION, "100").map(s -> s.getPath().toString()).collect(Collectors.toList());
    FileCreateUtils.createDeltaCommit(basePath, "100", fs);
    FileCreateUtils.createDeltaCommit(basePath, "101", fs);
    FileCreateUtils.createDeltaCommit(basePath, "102", fs);
    HoodieMergedLogRecordScanner scanner = HoodieMergedLogRecordScanner.newBuilder().withFileSystem(fs).withBasePath(basePath).withLogFilePaths(allLogFiles).withReaderSchema(schema).withLatestInstantTime("102").withMaxMemorySizeInBytes(10240L).withReadBlocksLazily(readBlocksLazily).withReverseReader(false).withBufferSize(bufferSize).withSpillableMapBasePath(BASE_OUTPUT_PATH).withDiskMapType(diskMapType).withBitCaskDiskMapCompressionEnabled(isCompressionEnabled).build();
    assertEquals(200, scanner.getTotalLogRecords(), "We still would read 200 records");
    final List<String> readKeys = new ArrayList<>(200);
    final List<Boolean> emptyPayloads = new ArrayList<>();
    scanner.forEach(s -> readKeys.add(s.getKey().getRecordKey()));
    scanner.forEach(s -> {
        try {
            if (!s.getData().getInsertValue(schema).isPresent()) {
                emptyPayloads.add(true);
            }
        } catch (IOException io) {
            throw new UncheckedIOException(io);
        }
    });
    assertEquals(200, readKeys.size(), "Stream collect should return all 200 records");
    assertEquals(50, emptyPayloads.size(), "Stream collect should return all 50 records with empty payloads");
    originalKeys.removeAll(deletedKeys);
    Collections.sort(originalKeys);
    Collections.sort(readKeys);
    assertEquals(originalKeys, readKeys, "CompositeAvroLogReader should return 150 records from 2 versions");
    // Rollback the last block
    header.put(HoodieLogBlock.HeaderMetadataType.INSTANT_TIME, "103");
    header.put(HoodieLogBlock.HeaderMetadataType.TARGET_INSTANT_TIME, "102");
    header.put(HoodieLogBlock.HeaderMetadataType.COMMAND_BLOCK_TYPE, String.valueOf(HoodieCommandBlock.HoodieCommandBlockTypeEnum.ROLLBACK_PREVIOUS_BLOCK.ordinal()));
    HoodieCommandBlock commandBlock = new HoodieCommandBlock(header);
    writer.appendBlock(commandBlock);
    FileCreateUtils.deleteDeltaCommit(basePath, "102", fs);
    readKeys.clear();
    scanner = HoodieMergedLogRecordScanner.newBuilder().withFileSystem(fs).withBasePath(basePath).withLogFilePaths(allLogFiles).withReaderSchema(schema).withLatestInstantTime("101").withMaxMemorySizeInBytes(10240L).withReadBlocksLazily(readBlocksLazily).withReverseReader(false).withBufferSize(bufferSize).withSpillableMapBasePath(BASE_OUTPUT_PATH).withDiskMapType(diskMapType).withBitCaskDiskMapCompressionEnabled(isCompressionEnabled).build();
    scanner.forEach(s -> readKeys.add(s.getKey().getRecordKey()));
    assertEquals(200, readKeys.size(), "Stream collect should return all 200 records after rollback of delete");
}

Example 7 with HoodieDeleteBlock

use of org.apache.hudi.common.table.log.block.HoodieDeleteBlock in project hudi by apache.

the class TestHoodieLogFormat method testAvroLogRecordReaderWithInsertsDeleteAndRollback.

@ParameterizedTest
@MethodSource("testArguments")
public void testAvroLogRecordReaderWithInsertsDeleteAndRollback(ExternalSpillableMap.DiskMapType diskMapType, boolean isCompressionEnabled, boolean readBlocksLazily) throws IOException, URISyntaxException, InterruptedException {
    // Write a 3 Data blocs with same InstantTime (written in same batch)
    Schema schema = HoodieAvroUtils.addMetadataFields(getSimpleSchema());
    // Set a small threshold so that every block is a new version
    Writer writer = HoodieLogFormat.newWriterBuilder().onParentPath(partitionPath).withFileExtension(HoodieLogFile.DELTA_EXTENSION).withFileId("test-fileid1").overBaseCommit("100").withFs(fs).build();
    // Write 1
    List<IndexedRecord> records1 = SchemaTestUtil.generateHoodieTestRecords(0, 100);
    List<IndexedRecord> copyOfRecords1 = records1.stream().map(record -> HoodieAvroUtils.rewriteRecord((GenericRecord) record, schema)).collect(Collectors.toList());
    Map<HoodieLogBlock.HeaderMetadataType, String> header = new HashMap<>();
    header.put(HoodieLogBlock.HeaderMetadataType.INSTANT_TIME, "100");
    header.put(HoodieLogBlock.HeaderMetadataType.TARGET_INSTANT_TIME, "100");
    header.put(HoodieLogBlock.HeaderMetadataType.SCHEMA, schema.toString());
    HoodieDataBlock dataBlock = getDataBlock(DEFAULT_DATA_BLOCK_TYPE, records1, header);
    writer.appendBlock(dataBlock);
    writer.appendBlock(dataBlock);
    writer.appendBlock(dataBlock);
    // Delete 50 keys
    // Delete 50 keys
    List<HoodieKey> deletedKeys = copyOfRecords1.stream().map(s -> (new HoodieKey(((GenericRecord) s).get(HoodieRecord.RECORD_KEY_METADATA_FIELD).toString(), ((GenericRecord) s).get(HoodieRecord.PARTITION_PATH_METADATA_FIELD).toString()))).collect(Collectors.toList()).subList(0, 50);
    HoodieDeleteBlock deleteBlock = new HoodieDeleteBlock(deletedKeys.toArray(new HoodieKey[50]), header);
    writer.appendBlock(deleteBlock);
    FileCreateUtils.createDeltaCommit(basePath, "100", fs);
    // Write 1 rollback block for a failed write
    header.put(HoodieLogBlock.HeaderMetadataType.INSTANT_TIME, "101");
    header.put(HeaderMetadataType.TARGET_INSTANT_TIME, "100");
    header.put(HoodieLogBlock.HeaderMetadataType.COMMAND_BLOCK_TYPE, String.valueOf(HoodieCommandBlock.HoodieCommandBlockTypeEnum.ROLLBACK_PREVIOUS_BLOCK.ordinal()));
    HoodieCommandBlock commandBlock = new HoodieCommandBlock(header);
    writer.appendBlock(commandBlock);
    List<String> allLogFiles = FSUtils.getAllLogFiles(fs, partitionPath, "test-fileid1", HoodieLogFile.DELTA_EXTENSION, "100").map(s -> s.getPath().toString()).collect(Collectors.toList());
    HoodieMergedLogRecordScanner scanner = HoodieMergedLogRecordScanner.newBuilder().withFileSystem(fs).withBasePath(basePath).withLogFilePaths(allLogFiles).withReaderSchema(schema).withLatestInstantTime("101").withMaxMemorySizeInBytes(10240L).withReadBlocksLazily(readBlocksLazily).withReverseReader(false).withBufferSize(bufferSize).withSpillableMapBasePath(BASE_OUTPUT_PATH).withDiskMapType(diskMapType).withBitCaskDiskMapCompressionEnabled(isCompressionEnabled).build();
    assertEquals(0, scanner.getTotalLogRecords(), "We would read 0 records");
}

Example 8 with HoodieDeleteBlock

use of org.apache.hudi.common.table.log.block.HoodieDeleteBlock in project hudi by apache.

the class TestHoodieLogFormat method testAvroLogRecordReaderWithFailedRollbacks.

@ParameterizedTest
@MethodSource("testArguments")
public void testAvroLogRecordReaderWithFailedRollbacks(ExternalSpillableMap.DiskMapType diskMapType, boolean isCompressionEnabled, boolean readBlocksLazily) throws IOException, URISyntaxException, InterruptedException {
    // Write a Data block and Delete block with same InstantTime (written in same batch)
    Schema schema = HoodieAvroUtils.addMetadataFields(getSimpleSchema());
    // Set a small threshold so that every block is a new version
    Writer writer = HoodieLogFormat.newWriterBuilder().onParentPath(partitionPath).withFileExtension(HoodieLogFile.DELTA_EXTENSION).withFileId("test-fileid1").overBaseCommit("100").withFs(fs).build();
    // Write 1
    List<IndexedRecord> records1 = SchemaTestUtil.generateHoodieTestRecords(0, 100);
    List<IndexedRecord> copyOfRecords1 = records1.stream().map(record -> HoodieAvroUtils.rewriteRecord((GenericRecord) record, schema)).collect(Collectors.toList());
    Map<HoodieLogBlock.HeaderMetadataType, String> header = new HashMap<>();
    header.put(HoodieLogBlock.HeaderMetadataType.INSTANT_TIME, "100");
    header.put(HoodieLogBlock.HeaderMetadataType.TARGET_INSTANT_TIME, "100");
    header.put(HoodieLogBlock.HeaderMetadataType.SCHEMA, schema.toString());
    HoodieDataBlock dataBlock = getDataBlock(DEFAULT_DATA_BLOCK_TYPE, records1, header);
    writer.appendBlock(dataBlock);
    // Write 2
    List<IndexedRecord> records2 = SchemaTestUtil.generateHoodieTestRecords(0, 100);
    header.put(HoodieLogBlock.HeaderMetadataType.SCHEMA, schema.toString());
    dataBlock = getDataBlock(DEFAULT_DATA_BLOCK_TYPE, records2, header);
    writer.appendBlock(dataBlock);
    // Delete 50 keys
    // Delete 50 keys
    List<HoodieKey> deletedKeys = copyOfRecords1.stream().map(s -> (new HoodieKey(((GenericRecord) s).get(HoodieRecord.RECORD_KEY_METADATA_FIELD).toString(), ((GenericRecord) s).get(HoodieRecord.PARTITION_PATH_METADATA_FIELD).toString()))).collect(Collectors.toList()).subList(0, 50);
    HoodieDeleteBlock deleteBlock = new HoodieDeleteBlock(deletedKeys.toArray(new HoodieKey[50]), header);
    writer.appendBlock(deleteBlock);
    FileCreateUtils.createDeltaCommit(basePath, "100", fs);
    // Attempt 1 : Write rollback block for a failed write
    header.put(HoodieLogBlock.HeaderMetadataType.COMMAND_BLOCK_TYPE, String.valueOf(HoodieCommandBlock.HoodieCommandBlockTypeEnum.ROLLBACK_PREVIOUS_BLOCK.ordinal()));
    HoodieCommandBlock commandBlock = new HoodieCommandBlock(header);
    try {
        writer.appendBlock(commandBlock);
        // Say job failed, retry writing 2 rollback in the next rollback(..) attempt
        throw new Exception("simulating failure");
    } catch (Exception e) {
    // it's okay
    }
    // Attempt 2 : Write another rollback blocks for a failed write
    writer.appendBlock(commandBlock);
    List<String> allLogFiles = FSUtils.getAllLogFiles(fs, partitionPath, "test-fileid1", HoodieLogFile.DELTA_EXTENSION, "100").map(s -> s.getPath().toString()).collect(Collectors.toList());
    // all data must be rolled back before merge
    HoodieMergedLogRecordScanner scanner = HoodieMergedLogRecordScanner.newBuilder().withFileSystem(fs).withBasePath(basePath).withLogFilePaths(allLogFiles).withReaderSchema(schema).withLatestInstantTime("100").withMaxMemorySizeInBytes(10240L).withReadBlocksLazily(readBlocksLazily).withReverseReader(false).withBufferSize(bufferSize).withSpillableMapBasePath(BASE_OUTPUT_PATH).withDiskMapType(diskMapType).withBitCaskDiskMapCompressionEnabled(isCompressionEnabled).build();
    assertEquals(0, scanner.getTotalLogRecords(), "We would have scanned 0 records because of rollback");
    final List<String> readKeys = new ArrayList<>();
    scanner.forEach(s -> readKeys.add(s.getKey().getRecordKey()));
    assertEquals(0, readKeys.size(), "Stream collect should return all 0 records");
    FileCreateUtils.deleteDeltaCommit(basePath, "100", fs);
}