Skip to content
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

More accurate estimate on parquet row groups size #11258

Open
wants to merge 8 commits into
base: main
Choose a base branch
from
+77 −9
Open

More accurate estimate on parquet row groups size #11258

Changes from 1 commit
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
07f6b97
add unit test for parquet row groups size
Oct 4, 2024
5854ccb
Estimate row group size
Oct 4, 2024
c83198e
code style
Oct 4, 2024
5e64668
fix testBinPackCombineMediumFiles test failure
Oct 7, 2024
742eaac
address comments
Oct 8, 2024
07fe927
remove -2 estimate
Oct 9, 2024
98ecfac
Merge branch 'main' of https://github.com/apache/iceberg into jinyang…
Oct 9, 2024
a6e9ef3
update unit tests
Oct 9, 2024
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
Prev Previous commit
Next Next commit
code style
  • Loading branch information
jinyang_li committed Oct 4, 2024
commit c83198e4319b67bd1282fc82a8277627df664cdb
36 changes: 18 additions & 18 deletions parquet/src/test/java/org/apache/iceberg/parquet/TestParquet.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -35,9 +35,8 @@
import java.nio.file.Path;
import java.util.Collections;
import java.util.List;
import java.util.Random;
import java.util.UUID;
import java.util.function.Function;
import java.util.stream.IntStream;
import org.apache.avro.generic.GenericData;
import org.apache.avro.generic.GenericRecord;
import org.apache.avro.generic.GenericRecordBuilder;
Expand Down Expand Up @@ -223,20 +222,18 @@ public void testTwoLevelList() throws IOException {
assertThat(recordRead.get("topbytes")).isEqualTo(expectedBinary);
}


@Test
public void testParquetRowGroupSize() throws IOException {
// verify parquet row group size should be close to configured size
ImmutableList.Builder<Types.NestedField> columnsBuilder = ImmutableList.builder();

for (int i = 1; i <= 50; i++) {
columnsBuilder.add(optional(i, "stringCol" + i, Types.StringType.get()));
}
int recordCount = 100000;
int columnCount = 50;

List<Types.NestedField> columns = columnsBuilder.build();
List<Types.NestedField> columns =
IntStream.rangeClosed(1, columnCount)
.mapToObj(i -> optional(i, "stringCol" + i, Types.StringType.get()))
.collect(ImmutableList.toImmutableList());
Schema schema = new Schema(columns);

int recordCount = 100000;
File file = createTempFile(temp);

List<GenericData.Record> records = Lists.newArrayListWithCapacity(recordCount);
Expand All @@ -252,19 +249,22 @@ public void testParquetRowGroupSize() throws IOException {
}

long actualSize =
write(
file,
schema,
ImmutableMap.of("write.parquet.row-group-size-bytes", "1048576"),
ParquetAvroWriter::buildWriter,
records.toArray(new GenericData.Record[] {}));
write(
file,
schema,
ImmutableMap.of("write.parquet.row-group-size-bytes", "1048576"),
ParquetAvroWriter::buildWriter,
records.toArray(new GenericData.Record[] {}));

try (ParquetFileReader reader = ParquetFileReader.open(ParquetIO.file(localInput(file)))) {
ParquetMetadata footer = reader.getFooter();
for (int i = 1; i < footer.getBlocks().size() - 1; i++) {
Copy link
Contributor

Choose a reason for hiding this comment

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

Suggested change
for (int i = 1; i < footer.getBlocks().size() - 1; i++) {
for (int i = 0; i < footer.getBlocks().size(); i++) {

Copy link
Author

Choose a reason for hiding this comment

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

the test skips first row group. The first row group is expected to be not accurate in size.

Copy link
Contributor

Choose a reason for hiding this comment

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

can you please add a comment so that this is clear for a reader why the first one is skipped?

BlockMetaData blockMetaData = footer.getBlocks().get(i);
System.out.println("Block " + i + " compressed size: " + blockMetaData.getCompressedSize());
assertThat(footer.getBlocks().get(i).getCompressedSize())
Copy link
Contributor

Choose a reason for hiding this comment

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

is it correct to look at the compressed size here instead of looking at getTotalByteSize()?

Copy link
Author

Choose a reason for hiding this comment

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

If parquet row group size is referring to on disk size, then it should be getCompressedSize().
in this test, blocks[i] compressed size is: 1,007,783, total size is: 41,462,625

.isBetween((long) 900 * 1024, (long) 1200 * 1024);
}

assertThat(footer.getBlocks().get(footer.getBlocks().size() - 1).getCompressedSize())
.isLessThan((long) 1200 * 1024);
}
}

Expand Down
Loading