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

Example 6 with HoodieCommandBlock

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

the class TestHoodieLogFormatAppendFailure method testFailedToGetAppendStreamFromHDFSNameNode.

@Test
@Timeout(60)
public void testFailedToGetAppendStreamFromHDFSNameNode() throws IOException, URISyntaxException, InterruptedException, TimeoutException {
    // Use some fs like LocalFileSystem, that does not support appends
    String uuid = UUID.randomUUID().toString();
    Path localPartitionPath = new Path("/tmp/");
    FileSystem fs = cluster.getFileSystem();
    Path testPath = new Path(localPartitionPath, uuid);
    fs.mkdirs(testPath);
    // Some data & append.
    List<IndexedRecord> records = SchemaTestUtil.generateTestRecords(0, 10);
    Map<HoodieLogBlock.HeaderMetadataType, String> header = new HashMap<>(2);
    header.put(HoodieLogBlock.HeaderMetadataType.INSTANT_TIME, "100");
    header.put(HoodieLogBlock.HeaderMetadataType.SCHEMA, getSimpleSchema().toString());
    HoodieAvroDataBlock dataBlock = new HoodieAvroDataBlock(records, header, HoodieRecord.RECORD_KEY_METADATA_FIELD);
    Writer writer = HoodieLogFormat.newWriterBuilder().onParentPath(testPath).withFileExtension(HoodieArchivedLogFile.ARCHIVE_EXTENSION).withFileId("commits.archive").overBaseCommit("").withFs(fs).build();
    writer.appendBlock(dataBlock);
    // get the current log file version to compare later
    int logFileVersion = writer.getLogFile().getLogVersion();
    Path logFilePath = writer.getLogFile().getPath();
    writer.close();
    // Wait for 3 times replication of file
    DFSTestUtil.waitReplication(fs, logFilePath, (short) 3);
    // Shut down all DNs that have the last block location for the file
    LocatedBlocks lbs = cluster.getFileSystem().getClient().getNamenode().getBlockLocations("/tmp/" + uuid + "/" + logFilePath.getName(), 0, Long.MAX_VALUE);
    List<DataNode> dnsOfCluster = cluster.getDataNodes();
    DatanodeInfo[] dnsWithLocations = lbs.getLastLocatedBlock().getLocations();
    for (DataNode dn : dnsOfCluster) {
        for (DatanodeInfo loc : dnsWithLocations) {
            if (dn.getDatanodeId().equals(loc)) {
                dn.shutdown();
                cluster.stopDataNode(dn.getDisplayName());
                DFSTestUtil.waitForDatanodeDeath(dn);
            }
        }
    }
    // Wait for the replication of this file to go down to 0
    DFSTestUtil.waitReplication(fs, logFilePath, (short) 0);
    // Opening a new Writer right now will throw IOException. The code should handle this, rollover the logfile and
    // return a new writer with a bumped up logVersion
    writer = HoodieLogFormat.newWriterBuilder().onParentPath(testPath).withFileExtension(HoodieArchivedLogFile.ARCHIVE_EXTENSION).withFileId("commits.archive").overBaseCommit("").withFs(fs).build();
    header = new HashMap<>();
    header.put(HoodieLogBlock.HeaderMetadataType.COMMAND_BLOCK_TYPE, String.valueOf(HoodieCommandBlock.HoodieCommandBlockTypeEnum.ROLLBACK_PREVIOUS_BLOCK.ordinal()));
    writer.appendBlock(new HoodieCommandBlock(header));
    // The log version should be different for this new writer
    assertNotEquals(writer.getLogFile().getLogVersion(), logFileVersion);
}

Example 7 with HoodieCommandBlock

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

the class BaseRollbackHelper method maybeDeleteAndCollectStats.

/**
 * May be delete interested files and collect stats or collect stats only.
 *
 * @param context           instance of {@link HoodieEngineContext} to use.
 * @param instantToRollback {@link HoodieInstant} of interest for which deletion or collect stats is requested.
 * @param rollbackRequests  List of {@link ListingBasedRollbackRequest} to be operated on.
 * @param doDelete          {@code true} if deletion has to be done. {@code false} if only stats are to be collected w/o performing any deletes.
 * @return stats collected with or w/o actual deletions.
 */
List<Pair<String, HoodieRollbackStat>> maybeDeleteAndCollectStats(HoodieEngineContext context, HoodieInstant instantToRollback, List<SerializableHoodieRollbackRequest> rollbackRequests, boolean doDelete, int numPartitions) {
    return context.flatMap(rollbackRequests, (SerializableFunction<SerializableHoodieRollbackRequest, Stream<Pair<String, HoodieRollbackStat>>>) rollbackRequest -> {
        List<String> filesToBeDeleted = rollbackRequest.getFilesToBeDeleted();
        if (!filesToBeDeleted.isEmpty()) {
            List<HoodieRollbackStat> rollbackStats = deleteFiles(metaClient, filesToBeDeleted, doDelete);
            List<Pair<String, HoodieRollbackStat>> partitionToRollbackStats = new ArrayList<>();
            rollbackStats.forEach(entry -> partitionToRollbackStats.add(Pair.of(entry.getPartitionPath(), entry)));
            return partitionToRollbackStats.stream();
        } else if (!rollbackRequest.getLogBlocksToBeDeleted().isEmpty()) {
            HoodieLogFormat.Writer writer = null;
            try {
                String fileId = rollbackRequest.getFileId();
                String latestBaseInstant = rollbackRequest.getLatestBaseInstant();
                writer = HoodieLogFormat.newWriterBuilder().onParentPath(FSUtils.getPartitionPath(metaClient.getBasePath(), rollbackRequest.getPartitionPath())).withFileId(fileId).overBaseCommit(latestBaseInstant).withFs(metaClient.getFs()).withFileExtension(HoodieLogFile.DELTA_EXTENSION).build();
                if (doDelete) {
                    Map<HoodieLogBlock.HeaderMetadataType, String> header = generateHeader(instantToRollback.getTimestamp());
                    writer.appendBlock(new HoodieCommandBlock(header));
                }
            } catch (IOException | InterruptedException io) {
                throw new HoodieRollbackException("Failed to rollback for instant " + instantToRollback, io);
            } finally {
                try {
                    if (writer != null) {
                        writer.close();
                    }
                } catch (IOException io) {
                    throw new HoodieIOException("Error appending rollback block", io);
                }
            }
            Map<FileStatus, Long> filesToNumBlocksRollback = Collections.singletonMap(metaClient.getFs().getFileStatus(Objects.requireNonNull(writer).getLogFile().getPath()), 1L);
            return Collections.singletonList(Pair.of(rollbackRequest.getPartitionPath(), HoodieRollbackStat.newBuilder().withPartitionPath(rollbackRequest.getPartitionPath()).withRollbackBlockAppendResults(filesToNumBlocksRollback).build())).stream();
        } else {
            return Collections.singletonList(Pair.of(rollbackRequest.getPartitionPath(), HoodieRollbackStat.newBuilder().withPartitionPath(rollbackRequest.getPartitionPath()).build())).stream();
        }
    }, numPartitions);
}

Example 8 with HoodieCommandBlock

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

the class HoodieLogFileReader method readBlock.

// TODO : convert content and block length to long by using ByteBuffer, raw byte [] allows
// for max of Integer size
private HoodieLogBlock readBlock() throws IOException {
    int blockSize;
    try {
        // 1 Read the total size of the block
        blockSize = (int) inputStream.readLong();
    } catch (EOFException | CorruptedLogFileException e) {
        // Create a corrupt block by finding the next MAGIC marker or EOF
        return createCorruptBlock();
    }
    // We may have had a crash which could have written this block partially
    // Skip blockSize in the stream and we should either find a sync marker (start of the next
    // block) or EOF. If we did not find either of it, then this block is a corrupted block.
    boolean isCorrupted = isBlockCorrupted(blockSize);
    if (isCorrupted) {
        return createCorruptBlock();
    }
    // 2. Read the version for this log format
    HoodieLogFormat.LogFormatVersion nextBlockVersion = readVersion();
    // 3. Read the block type for a log block
    HoodieLogBlockType blockType = tryReadBlockType(nextBlockVersion);
    // 4. Read the header for a log block, if present
    Map<HeaderMetadataType, String> header = nextBlockVersion.hasHeader() ? HoodieLogBlock.getLogMetadata(inputStream) : null;
    // 5. Read the content length for the content
    // Fallback to full-block size if no content-length
    // TODO replace w/ hasContentLength
    int contentLength = nextBlockVersion.getVersion() != HoodieLogFormatVersion.DEFAULT_VERSION ? (int) inputStream.readLong() : blockSize;
    // 6. Read the content or skip content based on IO vs Memory trade-off by client
    long contentPosition = inputStream.getPos();
    boolean shouldReadLazily = readBlockLazily && nextBlockVersion.getVersion() != HoodieLogFormatVersion.DEFAULT_VERSION;
    Option<byte[]> content = HoodieLogBlock.tryReadContent(inputStream, contentLength, shouldReadLazily);
    // 7. Read footer if any
    Map<HeaderMetadataType, String> footer = nextBlockVersion.hasFooter() ? HoodieLogBlock.getLogMetadata(inputStream) : null;
    // log file in reverse
    if (nextBlockVersion.hasLogBlockLength()) {
        inputStream.readLong();
    }
    // 9. Read the log block end position in the log file
    long blockEndPos = inputStream.getPos();
    HoodieLogBlock.HoodieLogBlockContentLocation logBlockContentLoc = new HoodieLogBlock.HoodieLogBlockContentLocation(hadoopConf, logFile, contentPosition, contentLength, blockEndPos);
    switch(Objects.requireNonNull(blockType)) {
        case AVRO_DATA_BLOCK:
            if (nextBlockVersion.getVersion() == HoodieLogFormatVersion.DEFAULT_VERSION) {
                return HoodieAvroDataBlock.getBlock(content.get(), readerSchema);
            } else {
                return new HoodieAvroDataBlock(inputStream, content, readBlockLazily, logBlockContentLoc, Option.ofNullable(readerSchema), header, footer, keyField);
            }
        case HFILE_DATA_BLOCK:
            checkState(nextBlockVersion.getVersion() != HoodieLogFormatVersion.DEFAULT_VERSION, String.format("HFile block could not be of version (%d)", HoodieLogFormatVersion.DEFAULT_VERSION));
            return new HoodieHFileDataBlock(inputStream, content, readBlockLazily, logBlockContentLoc, Option.ofNullable(readerSchema), header, footer, enableRecordLookups);
        case PARQUET_DATA_BLOCK:
            checkState(nextBlockVersion.getVersion() != HoodieLogFormatVersion.DEFAULT_VERSION, String.format("Parquet block could not be of version (%d)", HoodieLogFormatVersion.DEFAULT_VERSION));
            return new HoodieParquetDataBlock(inputStream, content, readBlockLazily, logBlockContentLoc, Option.ofNullable(readerSchema), header, footer, keyField);
        case DELETE_BLOCK:
            return new HoodieDeleteBlock(content, inputStream, readBlockLazily, Option.of(logBlockContentLoc), header, footer);
        case COMMAND_BLOCK:
            return new HoodieCommandBlock(content, inputStream, readBlockLazily, Option.of(logBlockContentLoc), header, footer);
        default:
            throw new HoodieNotSupportedException("Unsupported Block " + blockType);
    }
}

Example 9 with HoodieCommandBlock

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

the class TestHoodieLogFormat method testAvroLogRecordReaderWithMixedInsertsCorruptsAndRollback.

@ParameterizedTest
@MethodSource("testArguments")
public void testAvroLogRecordReaderWithMixedInsertsCorruptsAndRollback(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);
    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);
    writer.appendBlock(dataBlock);
    writer.appendBlock(dataBlock);
    writer.close();
    FileCreateUtils.createDeltaCommit(basePath, "100", fs);
    // Append some arbit byte[] to the end of the log (mimics a partially written commit)
    fs = FSUtils.getFs(fs.getUri().toString(), fs.getConf());
    FSDataOutputStream outputStream = fs.append(writer.getLogFile().getPath());
    // create a block with
    outputStream.write(HoodieLogFormat.MAGIC);
    outputStream.writeLong(1000);
    outputStream.writeInt(HoodieLogBlockType.AVRO_DATA_BLOCK.ordinal());
    outputStream.writeInt(HoodieLogFormat.CURRENT_VERSION);
    // Write out a length that does not confirm with the content
    outputStream.writeLong(100);
    outputStream.flush();
    outputStream.close();
    // Append some arbit byte[] to the end of the log (mimics a partially written commit)
    fs = FSUtils.getFs(fs.getUri().toString(), fs.getConf());
    outputStream = fs.append(writer.getLogFile().getPath());
    // create a block with
    outputStream.write(HoodieLogFormat.MAGIC);
    outputStream.writeLong(1000);
    outputStream.writeInt(HoodieLogBlockType.AVRO_DATA_BLOCK.ordinal());
    outputStream.writeInt(HoodieLogFormat.CURRENT_VERSION);
    // Write out a length that does not confirm with the content
    outputStream.writeLong(100);
    outputStream.flush();
    outputStream.close();
    writer = HoodieLogFormat.newWriterBuilder().onParentPath(partitionPath).withFileExtension(HoodieLogFile.DELTA_EXTENSION).withFileId("test-fileid1").overBaseCommit("100").withFs(fs).build();
    writer.appendBlock(dataBlock);
    writer.close();
    // Append some arbit byte[] to the end of the log (mimics a partially written commit)
    fs = FSUtils.getFs(fs.getUri().toString(), fs.getConf());
    outputStream = fs.append(writer.getLogFile().getPath());
    // create a block with
    outputStream.write(HoodieLogFormat.MAGIC);
    outputStream.writeLong(1000);
    outputStream.writeInt(HoodieLogBlockType.AVRO_DATA_BLOCK.ordinal());
    outputStream.writeInt(HoodieLogFormat.CURRENT_VERSION);
    // Write out a length that does not confirm with the content
    outputStream.writeLong(100);
    outputStream.flush();
    outputStream.close();
    writer = HoodieLogFormat.newWriterBuilder().onParentPath(partitionPath).withFileExtension(HoodieLogFile.DELTA_EXTENSION).withFileId("test-fileid1").overBaseCommit("100").withFs(fs).build();
    // Write 1 rollback block for the last commit instant
    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);
    writer.close();
    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");
    FileCreateUtils.deleteDeltaCommit(basePath, "100", fs);
}

Example 10 with HoodieCommandBlock

use of org.apache.hudi.common.table.log.block.HoodieCommandBlock 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");
}