Skip to content

Core: Support delete file stats in partitions metadata table #6661

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
szehon-ho merged 2 commits into from
May 5, 2023
Merged

Core: Support delete file stats in partitions metadata table #6661

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
974643b
Core: Support delete files stats in partitions metadata table
ajantha-bhat Jan 25, 2023
8f43f1e
Address nit
ajantha-bhat May 5, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
136 changes: 105 additions & 31 deletions core/src/main/java/org/apache/iceberg/PartitionsTable.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,9 @@

import com.github.benmanes.caffeine.cache.Caffeine;
import com.github.benmanes.caffeine.cache.LoadingCache;
import java.io.IOException;
import java.io.UncheckedIOException;
import java.util.List;
import org.apache.iceberg.expressions.ManifestEvaluator;
import org.apache.iceberg.io.CloseableIterable;
import org.apache.iceberg.relocated.com.google.common.annotations.VisibleForTesting;
Expand Down Expand Up @@ -47,7 +50,27 @@ public class PartitionsTable extends BaseMetadataTable {
Types.NestedField.required(
2, "record_count", Types.LongType.get(), "Count of records in data files"),
Types.NestedField.required(
3, "file_count", Types.IntegerType.get(), "Count of data files"));
3, "file_count", Types.IntegerType.get(), "Count of data files"),
Types.NestedField.required(
5,

@ajantha-bhat ajantha-bhat Mar 21, 2023

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Note:
spec_id (4) is present at the top.

"position_delete_record_count",
Types.LongType.get(),
"Count of records in position delete files"),
Types.NestedField.required(
6,
"position_delete_file_count",
Types.IntegerType.get(),
"Count of position delete files"),
Types.NestedField.required(
7,
"equality_delete_record_count",
Types.LongType.get(),
"Count of records in equality delete files"),
Types.NestedField.required(
8,
"equality_delete_file_count",
Types.IntegerType.get(),
"Count of equality delete files"));
}

@Override
Expand All @@ -58,7 +81,13 @@ public TableScan newScan() {
@Override
public Schema schema() {
if (table().spec().fields().size() < 1) {
return schema.select("record_count", "file_count");

@szehon-ho szehon-ho May 4, 2023

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think there's actually a bug here, and it only checks latest spec for Unpartitioned. if we have partition fields before but removed them, we will not show them. See other metadata tables like BaseFilesTable.schema.

Anyway its unrelated, but made #7533 to track it.

@ajantha-bhat ajantha-bhat May 5, 2023

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Good catch. I will explore this in the follow-up.

return schema.select(
"record_count",
"file_count",
"position_delete_record_count",
"position_delete_file_count",
"equality_delete_record_count",
"equality_delete_file_count");
}
return schema;
}
Expand All @@ -77,7 +106,14 @@ private DataTask task(StaticTableScan scan) {
schema(),
scan.schema(),
partitions,
root -> StaticDataTask.Row.of(root.dataRecordCount, root.dataFileCount));
root ->
StaticDataTask.Row.of(
root.dataRecordCount,
root.dataFileCount,
root.posDeleteRecordCount,
root.posDeleteFileCount,
root.eqDeleteRecordCount,
root.eqDeleteFileCount));
} else {
return StaticDataTask.of(
io().newInputFile(table().operations().current().metadataFileLocation()),
Expand All @@ -93,31 +129,55 @@ private static StaticDataTask.Row convertPartition(Partition partition) {
partition.partitionData,
partition.specId,
partition.dataRecordCount,
partition.dataFileCount);
partition.dataFileCount,
partition.posDeleteRecordCount,
partition.posDeleteFileCount,
partition.eqDeleteRecordCount,
partition.eqDeleteFileCount);
}

private static Iterable<Partition> partitions(Table table, StaticTableScan scan) {
Types.StructType partitionType = Partitioning.partitionType(table);
PartitionMap partitions = new PartitionMap(partitionType);

CloseableIterable<DataFile> datafiles = planDataFiles(scan);
for (DataFile dataFile : datafiles) {
StructLike partition =
PartitionUtil.coercePartition(
partitionType, table.specs().get(dataFile.specId()), dataFile.partition());
partitions.get(partition).update(dataFile);
try (CloseableIterable<ContentFile<?>> files = planFiles(scan)) {

@szehon-ho szehon-ho May 5, 2023

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for closing this

for (ContentFile<?> file : files) {
StructLike partition =
PartitionUtil.coercePartition(
partitionType, table.specs().get(file.specId()), file.partition());
partitions.get(partition).update(file);
}
} catch (IOException e) {
throw new UncheckedIOException(e);
}

return partitions.all();
}

@VisibleForTesting
static CloseableIterable<DataFile> planDataFiles(StaticTableScan scan) {
static CloseableIterable<ContentFile<?>> planFiles(StaticTableScan scan) {
Table table = scan.table();
Snapshot snapshot = scan.snapshot();

CloseableIterable<ManifestFile> dataManifests =
CloseableIterable.withNoopClose(snapshot.dataManifests(table.io()));
CloseableIterable<ManifestFile> filteredManifests =
filteredManifests(scan, table, scan.snapshot().allManifests(table.io()));

Iterable<CloseableIterable<ContentFile<?>>> tasks =
CloseableIterable.transform(
filteredManifests,
manifest ->
CloseableIterable.transform(
ManifestFiles.open(manifest, table.io(), table.specs())
.caseSensitive(scan.isCaseSensitive())
.select(BaseScan.DELETE_SCAN_COLUMNS), // don't select stats columns

@szehon-ho szehon-ho May 4, 2023 •
edited
Loading

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Nit: can do a switch on content to get either DELETE_SCAN_COLUMNS or SCAN_COLUMNS to make it clearer. Looks a bit weird, but I guess it works as DELETE_SCAN_COLUMNS is superset of SCAN_COLUMNS .

@ajantha-bhat ajantha-bhat May 5, 2023

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

updated it with the switch.

DELETE_SCAN_COLUMNS has only content type as an extra field. Which can work for both the data and the delete file.
Agree that the name looks weird when used for generic content. CONTENT_SCAN_COLUMNS would have been the more suitable name for it.

t -> (ContentFile<?>) t));

return new ParallelIterable<>(tasks, scan.planExecutor());
}

private static CloseableIterable<ManifestFile> filteredManifests(
StaticTableScan scan, Table table, List<ManifestFile> manifestFilesList) {
CloseableIterable<ManifestFile> manifestFiles =
CloseableIterable.withNoopClose(manifestFilesList);

LoadingCache<Integer, ManifestEvaluator> evalCache =
Caffeine.newBuilder()
Expand All @@ -129,19 +189,8 @@ static CloseableIterable<DataFile> planDataFiles(StaticTableScan scan) {
scan.filter(), transformedSpec, scan.isCaseSensitive());
});

CloseableIterable<ManifestFile> filteredManifests =
CloseableIterable.filter(
dataManifests, manifest -> evalCache.get(manifest.partitionSpecId()).eval(manifest));

Iterable<CloseableIterable<DataFile>> tasks =
CloseableIterable.transform(
filteredManifests,
manifest ->
ManifestFiles.read(manifest, table.io(), table.specs())
.caseSensitive(scan.isCaseSensitive())
.select(BaseScan.SCAN_COLUMNS)); // don't select stats columns

return new ParallelIterable<>(tasks, scan.planExecutor());
return CloseableIterable.filter(
manifestFiles, manifest -> evalCache.get(manifest.partitionSpecId()).eval(manifest));
}

private class PartitionsScan extends StaticTableScan {
Expand Down Expand Up @@ -182,18 +231,43 @@ static class Partition {
private int specId;
private long dataRecordCount;
private int dataFileCount;
private long posDeleteRecordCount;
private int posDeleteFileCount;
private long eqDeleteRecordCount;
private int eqDeleteFileCount;

Partition(StructLike key, Types.StructType keyType) {
this.partitionData = toPartitionData(key, keyType);
this.specId = 0;
this.dataRecordCount = 0;
this.dataFileCount = 0;
this.posDeleteRecordCount = 0;
this.posDeleteFileCount = 0;
this.eqDeleteRecordCount = 0;
this.eqDeleteFileCount = 0;
}

void update(DataFile file) {
this.dataRecordCount += file.recordCount();
this.dataFileCount += 1;
this.specId = file.specId();
void update(ContentFile<?> file) {
switch (file.content()) {
case DATA:
this.dataRecordCount += file.recordCount();
this.dataFileCount += 1;
this.specId = file.specId();
break;
case POSITION_DELETES:
this.posDeleteRecordCount = file.recordCount();
this.posDeleteFileCount += 1;
this.specId = file.specId();
break;

@szehon-ho szehon-ho May 3, 2023

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

How about specId here and below?

@ajantha-bhat ajantha-bhat May 4, 2023

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

For this partition value, while updating the data file count, the Spec id would have been updated.
I don't think there will be delete files without data file entries. So, I assumed that again updating here would be redundant. WDYT?

@ajantha-bhat ajantha-bhat May 4, 2023

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

updated it now thinking if the delete happens after the partition evolution, it should reflect the latest spec id.

case EQUALITY_DELETES:
this.eqDeleteRecordCount = file.recordCount();
this.eqDeleteFileCount += 1;
this.specId = file.specId();
break;
default:
throw new UnsupportedOperationException(
"Unsupported file content type: " + file.content());
}
}

/** Needed because StructProjection is not serializable */
Expand Down
12 changes: 6 additions & 6 deletions core/src/test/java/org/apache/iceberg/MetadataTableScanTestBase.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -81,18 +81,18 @@ protected void validateTaskScanResiduals(TableScan scan, boolean ignoreResiduals
}

protected void validateSingleFieldPartition(
CloseableIterable<DataFile> dataFiles, int partitionValue) {
validatePartition(dataFiles, 0, partitionValue);
CloseableIterable<ContentFile<?>> files, int partitionValue) {
validatePartition(files, 0, partitionValue);
}

protected void validatePartition(
CloseableIterable<DataFile> dataFiles, int position, int partitionValue) {
CloseableIterable<ContentFile<?>> files, int position, int partitionValue) {
Assert.assertTrue(
"File scan tasks do not include correct file",
StreamSupport.stream(dataFiles.spliterator(), false)
StreamSupport.stream(files.spliterator(), false)
.anyMatch(
dataFile -> {
StructLike partition = dataFile.partition();
file -> {
StructLike partition = file.partition();
if (position >= partition.size()) {
return false;
}
Expand Down
Loading