[HUDI-431] [WIP] Adding parquet data block with inline read support #3930
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,168 @@ | ||
| /* | ||
| * Licensed to the Apache Software Foundation (ASF) under one | ||
| * or more contributor license agreements. See the NOTICE file | ||
| * distributed with this work for additional information | ||
| * regarding copyright ownership. The ASF licenses this file | ||
| * to you under the Apache License, Version 2.0 (the | ||
| * "License"); you may not use this file except in compliance | ||
| * with the License. You may obtain a copy of the License at | ||
| * | ||
| * http://www.apache.org/licenses/LICENSE-2.0 | ||
| * | ||
| * Unless required by applicable law or agreed to in writing, software | ||
| * distributed under the License is distributed on an "AS IS" BASIS, | ||
| * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
| * See the License for the specific language governing permissions and | ||
| * limitations under the License. | ||
| */ | ||
|
|
||
| package org.apache.hudi.common.table.log.block; | ||
|
|
||
| import org.apache.hudi.avro.HoodieAvroWriteSupport; | ||
| import org.apache.hudi.common.bloom.BloomFilter; | ||
| import org.apache.hudi.common.bloom.BloomFilterFactory; | ||
| import org.apache.hudi.common.bloom.BloomFilterTypeCode; | ||
| import org.apache.hudi.common.fs.inline.InLineFSUtils; | ||
| import org.apache.hudi.common.fs.inline.InLineFileSystem; | ||
| import org.apache.hudi.common.model.HoodieLogFile; | ||
| import org.apache.hudi.common.model.HoodieRecord; | ||
| import org.apache.hudi.common.util.Option; | ||
| import org.apache.hudi.exception.HoodieIOException; | ||
| import org.apache.hudi.io.storage.HoodieAvroParquetConfig; | ||
| import org.apache.hudi.io.storage.HoodieParquetReader; | ||
| import org.apache.hudi.io.storage.HoodieParquetStreamWriter; | ||
|
|
||
| import org.apache.avro.Schema; | ||
| import org.apache.avro.generic.IndexedRecord; | ||
| import org.apache.hadoop.conf.Configuration; | ||
| import org.apache.hadoop.fs.FSDataInputStream; | ||
| import org.apache.hadoop.fs.Path; | ||
| import org.apache.parquet.avro.AvroSchemaConverter; | ||
| import org.apache.parquet.hadoop.ParquetWriter; | ||
| import org.apache.parquet.hadoop.metadata.CompressionCodecName; | ||
|
|
||
| import javax.annotation.Nonnull; | ||
|
|
||
| import java.io.ByteArrayOutputStream; | ||
| import java.io.DataOutputStream; | ||
| import java.io.IOException; | ||
| import java.util.ArrayList; | ||
| import java.util.HashMap; | ||
| import java.util.Iterator; | ||
| import java.util.List; | ||
| import java.util.Map; | ||
|
|
||
| /** | ||
| * HoodieParquetDataBlock contains a list of records serialized using Parquet. | ||
| */ | ||
| public class HoodieParquetDataBlock extends HoodieDataBlock { | ||
|
|
||
| public HoodieParquetDataBlock(@Nonnull Map<HeaderMetadataType, String> logBlockHeader, | ||
| @Nonnull Map<HeaderMetadataType, String> logBlockFooter, | ||
| @Nonnull Option<HoodieLogBlockContentLocation> blockContentLocation, @Nonnull Option<byte[]> content, | ||
| FSDataInputStream inputStream, boolean readBlockLazily) { | ||
| super(logBlockHeader, logBlockFooter, blockContentLocation, content, inputStream, readBlockLazily); | ||
| } | ||
|
|
||
| public HoodieParquetDataBlock(HoodieLogFile logFile, FSDataInputStream inputStream, Option<byte[]> content, | ||
| boolean readBlockLazily, long position, long blockSize, long blockEndpos, Schema readerSchema, | ||
| Map<HeaderMetadataType, String> header, Map<HeaderMetadataType, String> footer) { | ||
| super(content, inputStream, readBlockLazily, | ||
| Option.of(new HoodieLogBlockContentLocation(logFile, position, blockSize, blockEndpos)), readerSchema, header, | ||
| footer); | ||
| } | ||
|
|
||
| public HoodieParquetDataBlock(@Nonnull List<IndexedRecord> records, @Nonnull Map<HeaderMetadataType, String> header) { | ||
| super(records, header, new HashMap<>()); | ||
| } | ||
|
|
||
| @Override | ||
| public HoodieLogBlockType getBlockType() { | ||
| return HoodieLogBlockType.PARQUET_DATA_BLOCK; | ||
| } | ||
|
|
||
| @Override | ||
| protected byte[] serializeRecords() throws IOException { | ||
| // TODO: Need to decide from where to fetch all config values required below. We can't re-use index config as the purpose is different. | ||
|
Member
There was a problem hiding this comment. I think the configs here should match the configs for the dataset base file format. Under what case would the user want separate configs for base file and inline log blocks? |
||
| // And these are very specific to data blocks. Once we have consensus, we might need to route them to log block constructors ( | ||
| // as of now, log block constructors does not take in any configs in general). | ||
| BloomFilter filter = BloomFilterFactory.createBloomFilter( | ||
| Integer.parseInt("60000"),//HoodieIndexConfig.BLOOM_FILTER_NUM_ENTRIES.defaultValue()), | ||
| Double.parseDouble("0.000000001"),//HoodieIndexConfig.BLOOM_FILTER_FPP.defaultValue()), | ||
| Integer.parseInt("100000"),//HoodieIndexConfig.HOODIE_BLOOM_INDEX_FILTER_DYNAMIC_MAX_ENTRIES.defaultValue()), | ||
| BloomFilterTypeCode.SIMPLE.name());//HoodieIndexConfig.BLOOM_INDEX_FILTER_TYPE.defaultValue()); | ||
|
|
||
| HoodieAvroWriteSupport writeSupport = new HoodieAvroWriteSupport( | ||
| new AvroSchemaConverter().convert(schema), schema, Option.of(filter)); | ||
|
|
||
| HoodieAvroParquetConfig avroParquetConfig = new HoodieAvroParquetConfig(writeSupport, CompressionCodecName.GZIP, | ||
| ParquetWriter.DEFAULT_BLOCK_SIZE, ParquetWriter.DEFAULT_PAGE_SIZE, 1024 * 1024 * 1024, | ||
| new Configuration(), Double.parseDouble(String.valueOf(0.1)));//HoodieStorageConfig.PARQUET_COMPRESSION_RATIO.defaultValue())); | ||
|
|
||
| ByteArrayOutputStream baos = new ByteArrayOutputStream(); | ||
| DataOutputStream outputStream = null; | ||
| HoodieParquetStreamWriter<IndexedRecord> parquetWriter = null; | ||
| try { | ||
| outputStream = new DataOutputStream(baos); | ||
| parquetWriter = new HoodieParquetStreamWriter<>(outputStream, avroParquetConfig); | ||
| Iterator<IndexedRecord> itr = records.iterator(); | ||
| if (records.size() > 0) { | ||
| Schema.Field keyField = records.get(0).getSchema().getField(HoodieRecord.RECORD_KEY_METADATA_FIELD); | ||
| if (keyField == null) { | ||
| throw new HoodieIOException("Record key field missing from schema for records to be written to Parquet data block"); | ||
| } | ||
| while (itr.hasNext()) { | ||
| IndexedRecord record = itr.next(); | ||
| String recordKey = record.get(keyField.pos()).toString(); | ||
| parquetWriter.writeAvro(recordKey, record); | ||
| } | ||
| outputStream.flush(); | ||
| } | ||
| } finally { | ||
| if (outputStream != null) { | ||
| outputStream.close(); | ||
| } | ||
| if (parquetWriter != null) { | ||
| parquetWriter.close(); | ||
| } | ||
| } | ||
|
|
||
| return baos.toByteArray(); | ||
| } | ||
|
|
||
| @Override | ||
| public List<IndexedRecord> getRecords() { | ||
| try { | ||
| records = new ArrayList<>(); | ||
| // Get schema from the header | ||
| Schema writerSchema = new Schema.Parser().parse(super.getLogBlockHeader().get(HeaderMetadataType.SCHEMA)); | ||
| // If readerSchema was not present, use writerSchema | ||
| if (schema == null) { | ||
| schema = writerSchema; | ||
| } | ||
| Configuration conf = new Configuration(); | ||
| Configuration inlineConf = new Configuration(); | ||
| inlineConf.set("fs." + InLineFileSystem.SCHEME + ".impl", InLineFileSystem.class.getName()); | ||
|
|
||
| Path inlinePath = InLineFSUtils.getInlineFilePath( | ||
| getBlockContentLocation().get().getLogFile().getPath(), | ||
| getBlockContentLocation().get().getLogFile().getPath().getFileSystem(conf).getScheme(), | ||
| getBlockContentLocation().get().getContentPositionInLogFile(), | ||
| getBlockContentLocation().get().getBlockSize()); | ||
|
|
||
| HoodieParquetReader<IndexedRecord> parquetReader = new HoodieParquetReader<>(inlineConf, inlinePath); | ||
| Iterator<IndexedRecord> recordIterator = parquetReader.getRecordIterator(schema); | ||
| while (recordIterator.hasNext()) { | ||
| records.add(recordIterator.next()); | ||
| } | ||
| return records; | ||
| } catch (IOException e) { | ||
| throw new HoodieIOException("Reading parquet inlining failed ", e); | ||
| } | ||
| } | ||
|
|
||
| @Override | ||
| protected void deserializeRecords() throws IOException { | ||
| throw new IOException("Not implemented"); | ||
| } | ||
| } | ||
File renamed without changes.
File renamed without changes.
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think this config is not required.
For a dataset with a PARQUET file format, an inline log block format of PARQUET makes sense. Mix and match of base file format and inline log format does not seem to make writing data or query engine side faster.
So we can have a config - useInlineFiles() - when enabled, the inline file format will be format and it will have the same format as the base file format. This needs to be fixed for the dataset so can be from table config.