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

Example 1 with OrcDataSourceId

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

the class TestDwrfMetadataReader method testReadStripeFooterThrowsForLargeStreams.

@Test
public void testReadStripeFooterThrowsForLargeStreams() {
    DwrfProto.Stream stream = DwrfProto.Stream.newBuilder().setKind(DATA).setLength(Long.MAX_VALUE).build();
    DwrfProto.StripeFooter protoStripeFooter = DwrfProto.StripeFooter.newBuilder().addStreams(stream).build();
    byte[] data = protoStripeFooter.toByteArray();
    InputStream inputStream = new ByteArrayInputStream(data);
    OrcDataSourceId orcDataSourceId = new OrcDataSourceId("test");
    OrcCorruptionException ex = expectThrows(OrcCorruptionException.class, () -> dwrfMetadataReader.readStripeFooter(orcDataSourceId, ImmutableList.of(), inputStream));
    assertEquals(ex.getMessage(), "java.io.IOException: Malformed ORC file. Stream size 9223372036854775807 of one of the streams for column 0 is larger than supported size 2147483647 [test]");
}

Example 2 with OrcDataSourceId

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

the class SortingFileWriter method mergeFiles.

private void mergeFiles(Iterable<TempFile> files, Consumer<Page> consumer) {
    try (Closer closer = Closer.create()) {
        Collection<Iterator<Page>> iterators = new ArrayList<>();
        for (TempFile tempFile : files) {
            Path file = tempFile.getPath();
            OrcDataSource dataSource = new HdfsOrcDataSource(new OrcDataSourceId(file.toString()), fileSystem.getFileStatus(file).getLen(), new DataSize(1, MEGABYTE), new DataSize(8, MEGABYTE), new DataSize(8, MEGABYTE), false, fileSystem.open(file), new FileFormatDataSourceStats());
            closer.register(dataSource);
            iterators.add(new TempFileReader(types, dataSource));
        }
        new MergingPageIterator(iterators, types, sortFields, sortOrders).forEachRemaining(consumer);
        if (!sortedWriteToTempPathEnabled) {
            for (TempFile tempFile : files) {
                Path file = tempFile.getPath();
                fileSystem.delete(file, false);
                if (fileSystem.exists(file)) {
                    throw new IOException("Failed to delete temporary file: " + file);
                }
            }
        }
    } catch (IOException e) {
        throw new UncheckedIOException(e);
    }
}

Example 3 with OrcDataSourceId

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

the class StorageModule method createOrcFileTailSource.

@Singleton
@Provides
public OrcFileTailSource createOrcFileTailSource(OrcCacheConfig orcCacheConfig, MBeanExporter exporter) {
    int expectedFileTailSizeInBytes = toIntExact(orcCacheConfig.getExpectedFileTailSize().toBytes());
    boolean dwrfStripeCacheEnabled = orcCacheConfig.isDwrfStripeCacheEnabled();
    OrcFileTailSource orcFileTailSource = new StorageOrcFileTailSource(expectedFileTailSizeInBytes, dwrfStripeCacheEnabled);
    if (orcCacheConfig.isFileTailCacheEnabled()) {
        Cache<OrcDataSourceId, OrcFileTail> cache = CacheBuilder.newBuilder().maximumWeight(orcCacheConfig.getFileTailCacheSize().toBytes()).weigher((id, tail) -> ((OrcFileTail) tail).getTotalSize()).expireAfterAccess(orcCacheConfig.getFileTailCacheTtlSinceLastAccess().toMillis(), TimeUnit.MILLISECONDS).recordStats().build();
        CacheStatsMBean cacheStatsMBean = new CacheStatsMBean(cache);
        orcFileTailSource = new CachingOrcFileTailSource(orcFileTailSource, cache);
        exporter.export(generatedNameOf(CacheStatsMBean.class, connectorId + "_OrcFileTail"), cacheStatsMBean);
    }
    return orcFileTailSource;
}

Example 4 with OrcDataSourceId

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

the class HiveClientModule method createOrcFileTailSource.

@Singleton
@Provides
public OrcFileTailSource createOrcFileTailSource(OrcCacheConfig orcCacheConfig, MBeanExporter exporter) {
    int expectedFileTailSizeInBytes = toIntExact(orcCacheConfig.getExpectedFileTailSize().toBytes());
    boolean dwrfStripeCacheEnabled = orcCacheConfig.isDwrfStripeCacheEnabled();
    OrcFileTailSource orcFileTailSource = new StorageOrcFileTailSource(expectedFileTailSizeInBytes, dwrfStripeCacheEnabled);
    if (orcCacheConfig.isFileTailCacheEnabled()) {
        Cache<OrcDataSourceId, OrcFileTail> cache = CacheBuilder.newBuilder().maximumWeight(orcCacheConfig.getFileTailCacheSize().toBytes()).weigher((id, tail) -> ((OrcFileTail) tail).getFooterSize() + ((OrcFileTail) tail).getMetadataSize()).expireAfterAccess(orcCacheConfig.getFileTailCacheTtlSinceLastAccess().toMillis(), MILLISECONDS).recordStats().build();
        CacheStatsMBean cacheStatsMBean = new CacheStatsMBean(cache);
        orcFileTailSource = new CachingOrcFileTailSource(orcFileTailSource, cache);
        exporter.export(generatedNameOf(CacheStatsMBean.class, connectorId + "_OrcFileTail"), cacheStatsMBean);
    }
    return orcFileTailSource;
}

Example 5 with OrcDataSourceId

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

the class IcebergModule method createOrcFileTailSource.

@Singleton
@Provides
public OrcFileTailSource createOrcFileTailSource(OrcCacheConfig orcCacheConfig, MBeanExporter exporter) {
    int expectedFileTailSizeInBytes = toIntExact(orcCacheConfig.getExpectedFileTailSize().toBytes());
    boolean dwrfStripeCacheEnabled = orcCacheConfig.isDwrfStripeCacheEnabled();
    OrcFileTailSource orcFileTailSource = new StorageOrcFileTailSource(expectedFileTailSizeInBytes, dwrfStripeCacheEnabled);
    if (orcCacheConfig.isFileTailCacheEnabled()) {
        Cache<OrcDataSourceId, OrcFileTail> cache = CacheBuilder.newBuilder().maximumWeight(orcCacheConfig.getFileTailCacheSize().toBytes()).weigher((id, tail) -> ((OrcFileTail) tail).getFooterSize() + ((OrcFileTail) tail).getMetadataSize()).expireAfterAccess(orcCacheConfig.getFileTailCacheTtlSinceLastAccess().toMillis(), MILLISECONDS).recordStats().build();
        CacheStatsMBean cacheStatsMBean = new CacheStatsMBean(cache);
        orcFileTailSource = new CachingOrcFileTailSource(orcFileTailSource, cache);
        exporter.export(generatedNameOf(CacheStatsMBean.class, connectorId + "_OrcFileTail"), cacheStatsMBean);
    }
    return orcFileTailSource;
}