Skip to content

[WIP][VL] Add snappy to gluten columnar shuffle compression codec #11454

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
taiyang-li wants to merge 6 commits into
base: main
Choose a base branch
from
Open

[WIP][VL] Add snappy to gluten columnar shuffle compression codec #11454

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
2f6c02f
add snappy as gluten columnar shuffle compression codec
Jan 20, 2026
e667b97
[Velox] Add snappy support for columnar shuffle codec
taiyang-li Jan 24, 2026
c0f7d2c
[Velox] Add snappy support for columnar shuffle codec
taiyang-li Jan 24, 2026
f00b495
[Velox] Add snappy support for columnar shuffle codec
taiyang-li Jan 24, 2026
f1f8bf2
[Velox] Add snappy support for columnar shuffle codec
taiyang-li Jan 24, 2026
0213a52
[Velox] Add snappy support for columnar shuffle codec: implement SNAP…
Jan 24, 2026
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
2 changes: 1 addition & 1 deletion backends-velox/src/main/scala/org/apache/gluten/backendsapi/velox/VeloxBackend.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -91,7 +91,7 @@ object VeloxBackend {
}

object VeloxBackendSettings extends BackendSettingsApi {
val SHUFFLE_SUPPORTED_CODEC = Set("lz4", "zstd")
val SHUFFLE_SUPPORTED_CODEC = Set("lz4", "zstd", "snappy")
val GLUTEN_VELOX_UDF_LIB_PATHS = VeloxBackend.CONF_PREFIX + ".udfLibraryPaths"
val GLUTEN_VELOX_DRIVER_UDF_LIB_PATHS = VeloxBackend.CONF_PREFIX + ".driver.udfLibraryPaths"
val GLUTEN_VELOX_INTERNAL_UDF_LIB_PATHS = VeloxBackend.CONF_PREFIX + ".internal.udfLibraryPaths"
Expand Down
6 changes: 6 additions & 0 deletions cpp/core/utils/Compression.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -29,9 +29,15 @@ std::unique_ptr<arrow::util::Codec>
createCompressionCodec(arrow::Compression::type compressedType, CodecBackend codecBackend, int32_t compressionLevel) {
std::unique_ptr<arrow::util::Codec> codec;
switch (compressedType) {
case arrow::Compression::UNCOMPRESSED: {
return nullptr;
}
case arrow::Compression::LZ4_FRAME: {
GLUTEN_ASSIGN_OR_THROW(codec, arrow::util::Codec::Create(compressedType));
} break;
case arrow::Compression::SNAPPY: {
GLUTEN_ASSIGN_OR_THROW(codec, arrow::util::Codec::Create(compressedType));
} break;
case arrow::Compression::ZSTD: {
if (codecBackend == CodecBackend::NONE) {
GLUTEN_ASSIGN_OR_THROW(codec, arrow::util::Codec::Create(compressedType, compressionLevel));
Expand Down
5 changes: 4 additions & 1 deletion cpp/velox/tests/VeloxGpuShuffleWriterTest.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -137,7 +137,10 @@ std::vector<GpuShuffleTestParams> getTestParams() {
}

const std::vector<arrow::Compression::type> compressions = {
arrow::Compression::UNCOMPRESSED, arrow::Compression::LZ4_FRAME, arrow::Compression::ZSTD};
arrow::Compression::UNCOMPRESSED,
arrow::Compression::LZ4_FRAME,
arrow::Compression::ZSTD,
arrow::Compression::SNAPPY};
const std::vector<int32_t> compressionThresholds = {-1, 0, 3, 4, 10, 4096};
const std::vector<int32_t> mergeBufferSizes = {0, 3, 4, 10, 4096};

Expand Down
5 changes: 4 additions & 1 deletion cpp/velox/tests/VeloxShuffleWriterTest.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -99,7 +99,10 @@ std::vector<ShuffleTestParams> getTestParams() {
}

const std::vector<arrow::Compression::type> compressions = {
arrow::Compression::UNCOMPRESSED, arrow::Compression::LZ4_FRAME, arrow::Compression::ZSTD};
arrow::Compression::UNCOMPRESSED,
arrow::Compression::LZ4_FRAME,
arrow::Compression::ZSTD,
arrow::Compression::SNAPPY};
const std::vector<int32_t> compressionThresholds = {-1, 0, 3, 4, 10, 4096};
const std::vector<int32_t> mergeBufferSizes = {0, 3, 4, 10, 4096};

Expand Down
2 changes: 2 additions & 0 deletions gluten-substrait/src/main/scala/org/apache/spark/shuffle/GlutenShuffleUtils.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -105,6 +105,8 @@ object GlutenShuffleUtils {
checkAndGetBufferSize(IO_COMPRESSION_LZ4_BLOCKSIZE)
} else if ("zstd" == codec) {
checkAndGetBufferSize(IO_COMPRESSION_ZSTD_BUFFERSIZE)
} else if ("snappy" == codec) {
checkAndGetBufferSize(IO_COMPRESSION_SNAPPY_BLOCKSIZE)
} else if ("gzip" == codec) { // QAT supports it only.
// Temporarily hard-coded to 32k.
32 * 1024
Expand Down
Loading