Data: delete compaction optimization by bloom filter #5100
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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -19,6 +19,8 @@ | |
|
|
||
| package org.apache.iceberg.data; | ||
|
|
||
| import java.io.IOException; | ||
| import java.io.UncheckedIOException; | ||
| import java.util.Collection; | ||
| import java.util.List; | ||
| import java.util.Map; | ||
|
|
@@ -36,11 +38,15 @@ | |
| import org.apache.iceberg.data.parquet.GenericParquetReaders; | ||
| import org.apache.iceberg.deletes.Deletes; | ||
| import org.apache.iceberg.deletes.PositionDeleteIndex; | ||
| import org.apache.iceberg.expressions.Expression; | ||
| import org.apache.iceberg.expressions.Expressions; | ||
| import org.apache.iceberg.expressions.Literal; | ||
| import org.apache.iceberg.io.CloseableIterable; | ||
| import org.apache.iceberg.io.InputFile; | ||
| import org.apache.iceberg.orc.ORC; | ||
| import org.apache.iceberg.parquet.Parquet; | ||
| import org.apache.iceberg.parquet.ParquetBloomRowGroupFilter; | ||
| import org.apache.iceberg.parquet.ParquetUtil; | ||
| import org.apache.iceberg.relocated.com.google.common.base.Preconditions; | ||
| import org.apache.iceberg.relocated.com.google.common.collect.ImmutableList; | ||
| import org.apache.iceberg.relocated.com.google.common.collect.Iterables; | ||
|
|
@@ -49,10 +55,15 @@ | |
| import org.apache.iceberg.relocated.com.google.common.collect.Multimap; | ||
| import org.apache.iceberg.relocated.com.google.common.collect.Multimaps; | ||
| import org.apache.iceberg.relocated.com.google.common.collect.Sets; | ||
| import org.apache.iceberg.types.Type; | ||
| import org.apache.iceberg.types.TypeUtil; | ||
| import org.apache.iceberg.types.Types; | ||
| import org.apache.iceberg.util.StructLikeSet; | ||
| import org.apache.iceberg.util.StructProjection; | ||
| import org.apache.parquet.hadoop.BloomFilterReader; | ||
| import org.apache.parquet.hadoop.ParquetFileReader; | ||
| import org.apache.parquet.hadoop.metadata.BlockMetaData; | ||
| import org.apache.parquet.schema.MessageType; | ||
|
|
||
| public abstract class DeleteFilter<T> { | ||
| private static final long DEFAULT_SET_FILTER_THRESHOLD = 100_000L; | ||
|
|
@@ -133,10 +144,19 @@ public CloseableIterable<T> filter(CloseableIterable<T> records) { | |
|
|
||
| private List<Predicate<T>> applyEqDeletes() { | ||
| List<Predicate<T>> isInDeleteSets = Lists.newArrayList(); | ||
| Map<BlockMetaData, BloomFilterReader> parquetBloomFilterReader = Maps.newHashMap(); | ||
| ParquetFileReader parquetReader = null; | ||
| Predicate<Record> isInBloomFilter = null; | ||
| if (eqDeletes.isEmpty()) { | ||
| return isInDeleteSets; | ||
| } | ||
|
|
||
| // load bloomfilter readers from data file | ||
| if (filePath.endsWith(".parquet")) { | ||
| parquetReader = ParquetUtil.openFile(getInputFile(filePath)); | ||
|
Collaborator
There was a problem hiding this comment. Can we use
Author
There was a problem hiding this comment. Maybe it's a big change , I want to keep this reader open during the iteration of delete files, and considering orc format, the delete iteration need to be encapsulated in a new function, any advise for this change? |
||
| parquetBloomFilterReader.putAll(ParquetUtil.getParquetBloomFilters(parquetReader)); | ||
| } | ||
|
|
||
| Multimap<Set<Integer>, DeleteFile> filesByDeleteIds = Multimaps.newMultimap(Maps.newHashMap(), Lists::newArrayList); | ||
| for (DeleteFile delete : eqDeletes) { | ||
| filesByDeleteIds.put(Sets.newHashSet(delete.equalityFieldIds()), delete); | ||
|
|
@@ -148,6 +168,12 @@ private List<Predicate<T>> applyEqDeletes() { | |
|
|
||
| Schema deleteSchema = TypeUtil.select(requiredSchema, ids); | ||
|
|
||
| if (filePath.endsWith(".parquet") && parquetReader != null) { | ||
|
Collaborator
There was a problem hiding this comment. I think we can change the ctor parameter from
Author
There was a problem hiding this comment. yes, you are right, but dataFile is not passed into DeleteFilter as parameter, it was changed to filePath in #4381
Collaborator
There was a problem hiding this comment. I see, any concern if we change it to
Author
There was a problem hiding this comment. the change for constructor parameters is only for Trino supporting mor，Trino wrapped a dummy fileScanTask for data file currently, the author wants to remove fileScanTask implemented in Trino, and use the filePath parameter. If we change it to dataFile, compatibility is a problem. |
||
| MessageType fileSchema = parquetReader.getFileMetaData().getSchema(); | ||
| isInBloomFilter = | ||
| record -> findInParquetBloomFilter(record, deleteSchema, fileSchema, parquetBloomFilterReader); | ||
| } | ||
|
|
||
| // a projection to select and reorder fields of the file schema to match the delete rows | ||
| StructProjection projectRow = StructProjection.create(requiredSchema, deleteSchema); | ||
|
|
||
|
|
@@ -158,6 +184,11 @@ private List<Predicate<T>> applyEqDeletes() { | |
| CloseableIterable<Record> records = CloseableIterable.transform( | ||
| CloseableIterable.concat(deleteRecords), Record::copy); | ||
|
|
||
| // apply bloomfilter on delete records | ||
| if (isInBloomFilter != null) { | ||
| records = CloseableIterable.filter(records, isInBloomFilter); | ||
| } | ||
|
|
||
| StructLikeSet deleteSet = Deletes.toEqualitySet( | ||
| CloseableIterable.transform( | ||
| records, record -> new InternalRecordWrapper(deleteSchema.asStruct()).wrap(record)), | ||
|
|
@@ -166,10 +197,67 @@ private List<Predicate<T>> applyEqDeletes() { | |
| Predicate<T> isInDeleteSet = record -> deleteSet.contains(projectRow.wrap(asStructLike(record))); | ||
| isInDeleteSets.add(isInDeleteSet); | ||
| } | ||
| try { | ||
| if (parquetReader != null) { | ||
| parquetReader.close(); | ||
| } | ||
| } catch (IOException e) { | ||
| throw new UncheckedIOException("failed to close parquet file reader!", e); | ||
| } | ||
| return isInDeleteSets; | ||
| } | ||
|
|
||
| private boolean findInParquetBloomFilter( | ||
| Record record, | ||
| Schema deleteSchema, | ||
| MessageType fileSchema, | ||
| Map<BlockMetaData, BloomFilterReader> parquetBloomFilterReader) { | ||
| if (record.size() == 0 || parquetBloomFilterReader.isEmpty()) { | ||
| return true; | ||
| } | ||
| // build filter by record values | ||
| Expression filter = buildFilter(record, deleteSchema); | ||
| ParquetBloomRowGroupFilter bloomFilter = new ParquetBloomRowGroupFilter(deleteSchema, filter, true); | ||
| for (Map.Entry<BlockMetaData, BloomFilterReader> entry : parquetBloomFilterReader.entrySet()) { | ||
| boolean shouldRead = bloomFilter.shouldRead(fileSchema, entry.getKey(), entry.getValue()); | ||
| if (shouldRead) { | ||
| return true; | ||
| } | ||
| } | ||
|
|
||
| return false; | ||
| } | ||
|
|
||
| private Expression buildFilter(Record record, Schema schema) { | ||
| Expression filter = Expressions.alwaysTrue(); | ||
| for (Types.NestedField field : schema.columns()) { | ||
| Object value = getRecordValue(record, field); | ||
| if (value == null) { | ||
| continue; | ||
| } | ||
| filter = Expressions.and(filter, Expressions.equal(field.name(), value)); | ||
| } | ||
| return filter; | ||
| } | ||
|
|
||
| private Object getRecordValue(Record record, Types.NestedField field) { | ||
| Type type = field.type(); | ||
| switch (type.toString()) { | ||
| case "date": | ||
| return Literal.of(record.getField(field.name()).toString()).to(Types.DateType.get()).value(); | ||
| case "time": | ||
| return Literal.of(record.getField(field.name()).toString()).to(Types.TimeType.get()).value(); | ||
| case "timestamp": | ||
| if (((Types.TimestampType) type).shouldAdjustToUTC()) { | ||
| return Literal.of(record.getField(field.name()).toString()).to(Types.TimestampType.withZone()).value(); | ||
| } else { | ||
| return Literal.of(record.getField(field.name()).toString()).to(Types.TimestampType.withoutZone()).value(); | ||
| } | ||
| default: | ||
| return record.getField(field.name()); | ||
| } | ||
| } | ||
|
|
||
| public CloseableIterable<T> findEqualityDeleteRows(CloseableIterable<T> records) { | ||
| // Predicate to test whether a row has been deleted by equality deletions. | ||
| Predicate<T> deletedRows = applyEqDeletes().stream() | ||
|
|
||
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
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.
Do we want to check whether the bloom filter is turned on to avoid reading the footer if it is not?
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.
You mean we check the bloom filter by table properties, right? but the bloom filter properties may be updated, the bloom filter in current file is unmatched with table properties.