Skip to content

[Optimize] Optimize stripe footer multiple reads. #317

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
morningman merged 1 commit into from
May 22, 2025
Merged

[Optimize] Optimize stripe footer multiple reads. #317

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
9 changes: 2 additions & 7 deletions c++/src/Reader.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -679,7 +679,7 @@ namespace orc {
return std::unique_ptr<StripeInformation>(new StripeInformationImpl(
stripeInfo.offset(), stripeInfo.indexlength(), stripeInfo.datalength(),
stripeInfo.footerlength(), stripeInfo.numberofrows(), contents->stream.get(),
*contents->pool, contents->compression, contents->blockSize, contents->readerMetrics));
*contents->pool, contents->compression, contents->blockSize, contents->readerMetrics, nullptr));
}

FileVersion ReaderImpl::getFormatVersion() const {
Expand Down Expand Up @@ -1185,7 +1185,7 @@ namespace orc {
currentStripeInfo.offset(), currentStripeInfo.indexlength(),
currentStripeInfo.datalength(), currentStripeInfo.footerlength(),
currentStripeInfo.numberofrows(), contents->stream.get(), *contents->pool,
contents->compression, contents->blockSize, contents->readerMetrics));
contents->compression, contents->blockSize, contents->readerMetrics, &currentStripeFooter));
contents->stream->beforeReadStripe(std::move(currentStripeInformation), selectedColumns);

if (sargsApplier) {
Expand Down Expand Up @@ -1216,11 +1216,6 @@ namespace orc {

if (stringDictFilter != nullptr) {
std::list<std::string> dictFilterColumnNames;
std::unique_ptr<StripeInformation> currentStripeInformation(new StripeInformationImpl(
currentStripeInfo.offset(), currentStripeInfo.indexlength(),
currentStripeInfo.datalength(), currentStripeInfo.footerlength(),
currentStripeInfo.numberofrows(), contents->stream.get(), *contents->pool,
contents->compression, contents->blockSize, contents->readerMetrics));
stringDictFilter->fillDictFilterColumnNames(std::move(currentStripeInformation),
dictFilterColumnNames);
std::unordered_map<uint64_t, std::string> columnIdToNameMap;
Expand Down
5 changes: 3 additions & 2 deletions c++/src/StripeStream.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -130,13 +130,14 @@ namespace orc {
}

void StripeInformationImpl::ensureStripeFooterLoaded() const {
if (stripeFooter.get() == nullptr) {
if (stripeFooter == nullptr && managedStripeFooter.get() == nullptr) {
std::unique_ptr<SeekableInputStream> pbStream =
createDecompressor(compression,
std::make_unique<SeekableFileInputStream>(
stream, offset + indexLength + dataLength, footerLength, memory),
blockSize, memory, metrics);
stripeFooter = std::make_unique<proto::StripeFooter>();
managedStripeFooter = std::make_unique<proto::StripeFooter>();
stripeFooter = managedStripeFooter.get();
if (!stripeFooter->ParseFromZeroCopyStream(pbStream.get())) {
throw ParseError("Failed to parse the stripe footer");
}
Expand Down
8 changes: 5 additions & 3 deletions c++/src/StripeStream.hh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -132,15 +132,16 @@ namespace orc {
MemoryPool& memory;
CompressionKind compression;
uint64_t blockSize;
mutable std::unique_ptr<proto::StripeFooter> stripeFooter;
ReaderMetrics* metrics;
mutable proto::StripeFooter* stripeFooter;
mutable std::unique_ptr<proto::StripeFooter> managedStripeFooter;
void ensureStripeFooterLoaded() const;

public:
StripeInformationImpl(uint64_t _offset, uint64_t _indexLength, uint64_t _dataLength,
uint64_t _footerLength, uint64_t _numRows, InputStream* _stream,
MemoryPool& _memory, CompressionKind _compression, uint64_t _blockSize,
ReaderMetrics* _metrics)
ReaderMetrics* _metrics, proto::StripeFooter* _stripeFooter)
: offset(_offset),
indexLength(_indexLength),
dataLength(_dataLength),
Expand All @@ -150,7 +151,8 @@ namespace orc {
memory(_memory),
compression(_compression),
blockSize(_blockSize),
metrics(_metrics) {
metrics(_metrics),
stripeFooter(_stripeFooter) {
// PASS
}

Expand Down