Skip to content

GH-48111: [C++][Dataset] Avoid hang when fragment_readahead is 0 #55297

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

Open
555aaditya wants to merge 1 commit into
base: main
Choose a base branch
from
Open

GH-48111: [C++][Dataset] Avoid hang when fragment_readahead is 0 #55297

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
16 changes: 8 additions & 8 deletions cpp/src/arrow/dataset/scanner.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1030,14 +1030,14 @@ Result<acero::ExecNode*> MakeScanNode(acero::ExecPlan* plan,
FragmentsToBatches(std::move(fragment_gen), scan_options));

AsyncGenerator<EnumeratedRecordBatch> merged_batch_gen;
if (require_sequenced_output) {
if (scan_options->fragment_readahead > 1) {
ARROW_ASSIGN_OR_RAISE(merged_batch_gen, MakeSequencedMergedGenerator(
std::move(batch_gen_gen),
scan_options->fragment_readahead));
} else {
merged_batch_gen = MakeConcatenatedGenerator(std::move(batch_gen_gen));
}
if (scan_options->fragment_readahead <= 1) {
// No fragment readahead: scan one fragment at a time. MakeMergedGenerator
// would never subscribe to any fragment if fragment_readahead is 0.
merged_batch_gen = MakeConcatenatedGenerator(std::move(batch_gen_gen));
} else if (require_sequenced_output) {
ARROW_ASSIGN_OR_RAISE(merged_batch_gen,
MakeSequencedMergedGenerator(std::move(batch_gen_gen),
scan_options->fragment_readahead));
} else {
merged_batch_gen =
MakeMergedGenerator(std::move(batch_gen_gen), scan_options->fragment_readahead);
Expand Down
12 changes: 12 additions & 0 deletions cpp/src/arrow/dataset/scanner_test.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1049,6 +1049,18 @@ TEST_P(TestScanner, ScanBatchesUnordered) {
AssertScanBatchesUnorderedEqualRepetitionsOf(MakeScanner(batch), batch);
}

TEST_P(TestScanner, ScanWithoutFragmentReadahead) {
// GH-48111: unordered scans used to hang with fragment_readahead=0
SetSchema({field("i32", int32()), field("f64", float64())});
auto batch = ConstantArrayGenerator::Zeroes(GetParam().items_per_batch, schema_);
for (int fragment_readahead : {0, 1}) {
ARROW_SCOPED_TRACE("fragment_readahead = ", fragment_readahead);
options_->fragment_readahead = fragment_readahead;
AssertScanBatchesEqualRepetitionsOf(MakeScanner(batch), batch);
AssertScanBatchesUnorderedEqualRepetitionsOf(MakeScanner(batch), batch);
}
}

TEST_P(TestScanner, ScanWithCappedBatchSize) {
SetSchema({field("i32", int32()), field("f64", float64())});
auto batch = ConstantArrayGenerator::Zeroes(GetParam().items_per_batch, schema_);
Expand Down
8 changes: 8 additions & 0 deletions python/pyarrow/tests/test_dataset.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -565,6 +565,14 @@ def test_scanner_options(dataset):
assert batch.num_columns == 7


@pytest.mark.parquet
@pytest.mark.parametrize("fragment_readahead", [0, 1])
def test_scanner_no_fragment_readahead(dataset, fragment_readahead):
# GH-48111: this used to hang with fragment_readahead=0
table = dataset.scanner(fragment_readahead=fragment_readahead).to_table()
assert table.equals(dataset.to_table())


@pytest.mark.parquet
def test_scanner(dataset, dataset_reader):
scanner = dataset_reader.scanner(
Expand Down
Loading