Examples with OrcWriterStats com.facebook.presto.orc.OrcWriterStats used on opensource projects

Example 1 with OrcWriterStats

use of com.facebook.presto.orc.OrcWriterStats in project presto by prestodb.

the class TestWriterBlockRawSize method testFileMetadataRawSize.

@Test
public void testFileMetadataRawSize() throws IOException {
    Type type = INTEGER;
    List<Type> types = ImmutableList.of(type);
    int numBlocksPerRowGroup = 3;
    int numBlocksPerStripe = numBlocksPerRowGroup * 5;
    int numStripes = 4;
    int numBlocksPerFile = numBlocksPerStripe * numStripes + 1;
    BlockBuilder blockBuilder = type.createBlockBuilder(null, NUM_ELEMENTS * 2);
    for (int i = 0; i < NUM_ELEMENTS; i++) {
        blockBuilder.appendNull();
        type.writeLong(blockBuilder, i);
    }
    long blockRawSize = ((FixedWidthType) type).getFixedSize() * NUM_ELEMENTS + NUM_ELEMENTS;
    Block block = blockBuilder.build();
    Block[] blocks = new Block[] { block };
    OrcWriterOptions writerOptions = OrcWriterOptions.builder().withRowGroupMaxRowCount(block.getPositionCount() * numBlocksPerRowGroup).withFlushPolicy(DefaultOrcWriterFlushPolicy.builder().withStripeMaxRowCount(block.getPositionCount() * numBlocksPerStripe).build()).build();
    for (OrcEncoding encoding : OrcEncoding.values()) {
        try (TempFile tempFile = new TempFile()) {
            OrcWriter writer = createOrcWriter(tempFile.getFile(), encoding, ZSTD, Optional.empty(), types, writerOptions, new OrcWriterStats());
            for (int i = 0; i < numBlocksPerFile; i++) {
                writer.write(new Page(blocks));
            }
            writer.close();
            writer.validate(new FileOrcDataSource(tempFile.getFile(), new DataSize(1, MEGABYTE), new DataSize(1, MEGABYTE), new DataSize(1, MEGABYTE), true));
            Footer footer = OrcTester.getFileMetadata(tempFile.getFile(), encoding).getFooter();
            verifyValue(encoding, footer.getRawSize(), blockRawSize * numBlocksPerFile);
            assertEquals(footer.getStripes().size(), numStripes + 1);
            int numBlocksRemaining = numBlocksPerFile;
            for (StripeInformation stripeInfo : footer.getStripes()) {
                int numBlocksInStripe = Math.min(numBlocksRemaining, numBlocksPerStripe);
                verifyValue(encoding, stripeInfo.getRawDataSize(), blockRawSize * numBlocksInStripe);
                numBlocksRemaining -= numBlocksInStripe;
            }
        }
    }
}

Example 2 with OrcWriterStats

use of com.facebook.presto.orc.OrcWriterStats in project presto by prestodb.

the class IcebergFileWriterFactory method createOrcWriter.

private IcebergFileWriter createOrcWriter(Path outputPath, Schema icebergSchema, JobConf jobConf, ConnectorSession session) {
    try {
        FileSystem fileSystem = hdfsEnvironment.getFileSystem(session.getUser(), outputPath, jobConf);
        DataSink orcDataSink = hdfsEnvironment.doAs(session.getUser(), () -> new OutputStreamDataSink(fileSystem.create(outputPath)));
        Callable<Void> rollbackAction = () -> {
            hdfsEnvironment.doAs(session.getUser(), () -> fileSystem.delete(outputPath, false));
            return null;
        };
        List<Types.NestedField> columnFields = icebergSchema.columns();
        List<String> fileColumnNames = columnFields.stream().map(Types.NestedField::name).collect(toImmutableList());
        List<Type> fileColumnTypes = columnFields.stream().map(Types.NestedField::type).map(type -> toPrestoType(type, typeManager)).collect(toImmutableList());
        Optional<Supplier<OrcDataSource>> validationInputFactory = Optional.empty();
        if (isOrcOptimizedWriterValidate(session)) {
            validationInputFactory = Optional.of(() -> {
                try {
                    return new HdfsOrcDataSource(new OrcDataSourceId(outputPath.toString()), hdfsEnvironment.doAs(session.getUser(), () -> fileSystem.getFileStatus(outputPath).getLen()), getOrcMaxMergeDistance(session), getOrcMaxBufferSize(session), getOrcStreamBufferSize(session), false, hdfsEnvironment.doAs(session.getUser(), () -> fileSystem.open(outputPath)), readStats);
                } catch (IOException e) {
                    throw new PrestoException(ICEBERG_WRITE_VALIDATION_FAILED, e);
                }
            });
        }
        return new IcebergOrcFileWriter(icebergSchema, orcDataSink, rollbackAction, ORC, fileColumnNames, fileColumnTypes, toOrcType(icebergSchema), getCompressionCodec(session).getOrcCompressionKind(), orcFileWriterConfig.toOrcWriterOptionsBuilder().withFlushPolicy(DefaultOrcWriterFlushPolicy.builder().withStripeMinSize(HiveSessionProperties.getOrcOptimizedWriterMinStripeSize(session)).withStripeMaxSize(HiveSessionProperties.getOrcOptimizedWriterMaxStripeSize(session)).withStripeMaxRowCount(HiveSessionProperties.getOrcOptimizedWriterMaxStripeRows(session)).build()).withDictionaryMaxMemory(HiveSessionProperties.getOrcOptimizedWriterMaxDictionaryMemory(session)).withMaxStringStatisticsLimit(HiveSessionProperties.getOrcStringStatisticsLimit(session)).build(), IntStream.range(0, fileColumnNames.size()).toArray(), ImmutableMap.<String, String>builder().put(PRESTO_VERSION_NAME, nodeVersion.toString()).put(PRESTO_QUERY_ID_NAME, session.getQueryId()).build(), UTC, validationInputFactory, getOrcOptimizedWriterValidateMode(session), orcWriterStats, dwrfEncryptionProvider, Optional.empty());
    } catch (IOException e) {
        throw new PrestoException(ICEBERG_WRITER_OPEN_ERROR, "Error creating ORC file", e);
    }
}