Logo ROOT  
Reference Guide
 
Loading...
Searching...
No Matches
ROOT::Experimental::Internal::RPageSinkDaos Class Reference

Storage provider that writes ntuple pages to into a DAOS container.

Currently, an object is allocated for ntuple metadata (anchor/header/footer). Objects can correspond to pages or clusters of pages depending on the RNTuple-DAOS mapping strategy.

Definition at line 101 of file RPageStorageDaos.hxx.

Public Member Functions

 RPageSinkDaos (std::string_view ntupleName, std::string_view uri, const RNTupleWriteOptions &options)
 
 ~RPageSinkDaos () override
 
- Public Member Functions inherited from ROOT::Experimental::Internal::RPagePersistentSink
 RPagePersistentSink (const RPagePersistentSink &)=delete
 
 RPagePersistentSink (RPagePersistentSink &&)=default
 
 RPagePersistentSink (std::string_view ntupleName, const RNTupleWriteOptions &options)
 
 ~RPagePersistentSink () override
 
ColumnHandle_t AddColumn (DescriptorId_t fieldId, RColumn &column) final
 Register a new column.
 
void CommitClusterGroup () final
 Write out the page locations (page list envelope) for all the committed clusters since the last call of CommitClusterGroup (or the beginning of writing).
 
void CommitPage (ColumnHandle_t columnHandle, const RPage &page) final
 Write a page to the storage. The column must have been added before.
 
void CommitSealedPage (DescriptorId_t physicalColumnId, const RPageStorage::RSealedPage &sealedPage) final
 Write a preprocessed page to storage. The column must have been added before.
 
void CommitSealedPageV (std::span< RPageStorage::RSealedPageGroup > ranges) final
 Write a vector of preprocessed pages to storage. The corresponding columns must have been added before.
 
void CommitStagedClusters (std::span< RStagedCluster > clusters) final
 Commit staged clusters, logically appending them to the ntuple descriptor.
 
void CommitSuppressedColumn (ColumnHandle_t columnHandle) final
 Commits a suppressed column for the current cluster.
 
const RNTupleDescriptorGetDescriptor () const final
 Return the RNTupleDescriptor being constructed.
 
void InitFromDescriptor (const RNTupleDescriptor &descriptor)
 Initialize sink based on an existing descriptor and fill into the descriptor builder.
 
RPagePersistentSinkoperator= (const RPagePersistentSink &)=delete
 
RPagePersistentSinkoperator= (RPagePersistentSink &&)=default
 
RStagedCluster StageCluster (NTupleSize_t nNewEntries) final
 Stage the current cluster and create a new one for the following data.
 
void UpdateExtraTypeInfo (const RExtraTypeInfoDescriptor &extraTypeInfo) final
 Adds an extra type information record to schema.
 
void UpdateSchema (const RNTupleModelChangeset &changeset, NTupleSize_t firstEntry) final
 Incorporate incremental changes to the model into the ntuple descriptor.
 
- Public Member Functions inherited from ROOT::Experimental::Internal::RPageSink
 RPageSink (const RPageSink &)=delete
 
 RPageSink (RPageSink &&)=default
 
 RPageSink (std::string_view ntupleName, const RNTupleWriteOptions &options)
 
 ~RPageSink () override
 
virtual std::uint64_t CommitCluster (NTupleSize_t nNewEntries)
 Finalize the current cluster and create a new one for the following data.
 
void CommitDataset ()
 Run the registered callbacks and finalize the current cluster and the entrire data set.
 
void DropColumn (ColumnHandle_t) final
 Unregisters a column.
 
virtual RSinkGuard GetSinkGuard ()
 
EPageStorageType GetType () final
 Whether the concrete implementation is a sink or a source.
 
const RNTupleWriteOptionsGetWriteOptions () const
 Returns the sink's write options.
 
void Init (RNTupleModel &model)
 Physically creates the storage container to hold the ntuple (e.g., a keys a TFile or an S3 bucket) Init() associates column handles to the columns referenced by the model.
 
bool IsInitialized () const
 
RPageSinkoperator= (const RPageSink &)=delete
 
RPageSinkoperator= (RPageSink &&)=default
 
void RegisterOnCommitDatasetCallback (Callback_t callback)
 The registered callback is executed at the beginning of CommitDataset();.
 
virtual RPage ReservePage (ColumnHandle_t columnHandle, std::size_t nElements)
 Get a new, empty page for the given column that can be filled with up to nElements; nElements must be larger than zero.
 
- Public Member Functions inherited from ROOT::Experimental::Internal::RPageStorage
 RPageStorage (const RPageStorage &other)=delete
 
 RPageStorage (RPageStorage &&other)=default
 
 RPageStorage (std::string_view name)
 
virtual ~RPageStorage ()
 
ColumnId_t GetColumnId (ColumnHandle_t columnHandle) const
 
virtual Detail::RNTupleMetricsGetMetrics ()
 Returns the default metrics object.
 
const std::string & GetNTupleName () const
 Returns the NTuple name.
 
RPageStorageoperator= (const RPageStorage &other)=delete
 
RPageStorageoperator= (RPageStorage &&other)=default
 
void SetTaskScheduler (RTaskScheduler *taskScheduler)
 

Protected Member Functions

RNTupleLocator CommitClusterGroupImpl (unsigned char *serializedPageList, std::uint32_t length) final
 Returns the locator of the page list envelope of the given buffer that contains the serialized page list.
 
void CommitDatasetImpl () final
 
void CommitDatasetImpl (unsigned char *serializedFooter, std::uint32_t length) final
 
virtual void CommitDatasetImpl (unsigned char *serializedFooter, std::uint32_t length)=0
 
RNTupleLocator CommitPageImpl (ColumnHandle_t columnHandle, const RPage &page) final
 
RNTupleLocator CommitSealedPageImpl (DescriptorId_t physicalColumnId, const RPageStorage::RSealedPage &sealedPage) final
 
std::vector< RNTupleLocatorCommitSealedPageVImpl (std::span< RPageStorage::RSealedPageGroup > ranges, const std::vector< bool > &mask) final
 Vector commit of preprocessed pages.
 
void InitImpl (RNTupleModel &model) final
 Updates the descriptor and calls InitImpl() that handles the backend-specific details (file, DAOS, etc.)
 
void InitImpl (unsigned char *serializedHeader, std::uint32_t length) final
 
virtual void InitImpl (unsigned char *serializedHeader, std::uint32_t length)=0
 
std::uint64_t StageClusterImpl () final
 Returns the number of bytes written to storage (excluding metadata)
 
void WriteNTupleAnchor ()
 
void WriteNTupleFooter (const void *data, size_t nbytes, size_t lenFooter)
 
void WriteNTupleHeader (const void *data, size_t nbytes, size_t lenHeader)
 
- Protected Member Functions inherited from ROOT::Experimental::Internal::RPagePersistentSink
void EnableDefaultMetrics (const std::string &prefix)
 Enables the default set of metrics provided by RPageSink.
 
- Protected Member Functions inherited from ROOT::Experimental::Internal::RPageSink
RSealedPage SealPage (const RPage &page, const RColumnElementBase &element)
 Helper for streaming a page.
 
- Protected Member Functions inherited from ROOT::Experimental::Internal::RPageStorage
void WaitForAllTasks ()
 

Private Attributes

uint32_t fCageSizeLimit {}
 
std::atomic< std::uint64_t > fClusterGroupId {0}
 Cluster group counter for the next committed cluster pagelist; incremented in CommitClusterGroupImpl()
 
std::unique_ptr< RDaosContainerfDaosContainer
 Underlying DAOS container.
 
std::uint64_t fNBytesCurrentCluster {0}
 Tracks the number of bytes committed to the current cluster.
 
RDaosNTupleAnchor fNTupleAnchor
 
ntuple_index_t fNTupleIndex {0}
 
std::atomic< std::uint64_t > fPageId {0}
 Page identifier for the next committed page; it is automatically incremented in CommitSealedPageImpl()
 
std::string fURI
 A URI to a DAOS pool of the form 'daos://pool-label/container-label'.
 

Additional Inherited Members

- Public Types inherited from ROOT::Experimental::Internal::RPageSink
using Callback_t = std::function< void(RPageSink &)>
 
- Public Types inherited from ROOT::Experimental::Internal::RPageStorage
using ColumnHandle_t = RColumnHandle
 The column handle identifies a column with the current open page storage.
 
using SealedPageSequence_t = std::deque< RSealedPage >
 
- Static Public Member Functions inherited from ROOT::Experimental::Internal::RPagePersistentSink
static std::unique_ptr< RPageSinkCreate (std::string_view ntupleName, std::string_view location, const RNTupleWriteOptions &options=RNTupleWriteOptions())
 Guess the concrete derived page source from the location.
 
- Static Public Member Functions inherited from ROOT::Experimental::Internal::RPageSink
static RSealedPage SealPage (const RSealPageConfig &config)
 Seal a page using the provided info.
 
- Static Public Attributes inherited from ROOT::Experimental::Internal::RPageStorage
static constexpr std::size_t kNBytesPageChecksum = sizeof(std::uint64_t)
 The page checksum is a 64bit xxhash3.
 
- Protected Attributes inherited from ROOT::Experimental::Internal::RPagePersistentSink
std::unique_ptr< RCountersfCounters
 
Internal::RNTupleDescriptorBuilder fDescriptorBuilder
 
RFeatures fFeatures
 
- Protected Attributes inherited from ROOT::Experimental::Internal::RPageSink
std::unique_ptr< RNTupleCompressorfCompressor
 Helper to zip pages and header/footer; includes a 16MB (kMAXZIPBUF) zip buffer.
 
std::unique_ptr< RNTupleWriteOptionsfOptions
 
- Protected Attributes inherited from ROOT::Experimental::Internal::RPageStorage
Detail::RNTupleMetrics fMetrics
 
std::string fNTupleName
 
std::unique_ptr< RPageAllocatorfPageAllocator
 For the time being, we will use the heap allocator for all sources and sinks. This may change in the future.
 
RTaskSchedulerfTaskScheduler = nullptr
 

#include <ROOT/RPageStorageDaos.hxx>

Inheritance diagram for ROOT::Experimental::Internal::RPageSinkDaos:
[legend]

Constructor & Destructor Documentation

◆ RPageSinkDaos()

ROOT::Experimental::Internal::RPageSinkDaos::RPageSinkDaos ( std::string_view  ntupleName,
std::string_view  uri,
const RNTupleWriteOptions options 
)

Definition at line 283 of file RPageStorageDaos.cxx.

◆ ~RPageSinkDaos()

ROOT::Experimental::Internal::RPageSinkDaos::~RPageSinkDaos ( )
overridedefault

Member Function Documentation

◆ CommitClusterGroupImpl()

ROOT::Experimental::RNTupleLocator ROOT::Experimental::Internal::RPageSinkDaos::CommitClusterGroupImpl ( unsigned char *  serializedPageList,
std::uint32_t  length 
)
finalprotectedvirtual

Returns the locator of the page list envelope of the given buffer that contains the serialized page list.

Typically, the implementation takes care of compressing and writing the provided buffer.

Implements ROOT::Experimental::Internal::RPagePersistentSink.

Definition at line 437 of file RPageStorageDaos.cxx.

◆ CommitDatasetImpl() [1/3]

void ROOT::Experimental::Internal::RPagePersistentSink::CommitDatasetImpl ( )
finalprotectedvirtual

Reimplemented from ROOT::Experimental::Internal::RPagePersistentSink.

Definition at line 535 of file RPageStorage.cxx.

◆ CommitDatasetImpl() [2/3]

void ROOT::Experimental::Internal::RPageSinkDaos::CommitDatasetImpl ( unsigned char *  serializedFooter,
std::uint32_t  length 
)
finalprotectedvirtual

◆ CommitDatasetImpl() [3/3]

virtual void ROOT::Experimental::Internal::RPagePersistentSink::CommitDatasetImpl ( unsigned char *  serializedFooter,
std::uint32_t  length 
)
protectedvirtual

◆ CommitPageImpl()

ROOT::Experimental::RNTupleLocator ROOT::Experimental::Internal::RPageSinkDaos::CommitPageImpl ( ColumnHandle_t  columnHandle,
const RPage page 
)
finalprotectedvirtual

◆ CommitSealedPageImpl()

ROOT::Experimental::RNTupleLocator ROOT::Experimental::Internal::RPageSinkDaos::CommitSealedPageImpl ( DescriptorId_t  physicalColumnId,
const RPageStorage::RSealedPage sealedPage 
)
finalprotectedvirtual

◆ CommitSealedPageVImpl()

std::vector< ROOT::Experimental::RNTupleLocator > ROOT::Experimental::Internal::RPageSinkDaos::CommitSealedPageVImpl ( std::span< RPageStorage::RSealedPageGroup ranges,
const std::vector< bool > &  mask 
)
finalprotectedvirtual

Vector commit of preprocessed pages.

The ranges array specifies a range of sealed pages to be committed for each column. The returned vector contains, in order, the RNTupleLocator for each page on each range in ranges, i.e. the first N entries refer to the N pages in ranges[0], followed by M entries that refer to the M pages in ranges[1], etc. The mask allows to skip writing out certain pages. The vector has the size of all the pages. For every false value in the mask, the corresponding locator is skipped (missing) in the output vector. The default is to call CommitSealedPageImpl for each page; derived classes may provide an optimized implementation though.

Reimplemented from ROOT::Experimental::Internal::RPagePersistentSink.

Definition at line 364 of file RPageStorageDaos.cxx.

◆ InitImpl() [1/3]

void ROOT::Experimental::Internal::RPagePersistentSink::InitImpl ( RNTupleModel model)
finalprotectedvirtual

Updates the descriptor and calls InitImpl() that handles the backend-specific details (file, DAOS, etc.)

Reimplemented from ROOT::Experimental::Internal::RPagePersistentSink.

Definition at line 521 of file RPageStorage.cxx.

◆ InitImpl() [2/3]

void ROOT::Experimental::Internal::RPageSinkDaos::InitImpl ( unsigned char *  serializedHeader,
std::uint32_t  length 
)
finalprotectedvirtual

◆ InitImpl() [3/3]

virtual void ROOT::Experimental::Internal::RPagePersistentSink::InitImpl ( unsigned char *  serializedHeader,
std::uint32_t  length 
)
protectedvirtual

◆ StageClusterImpl()

std::uint64_t ROOT::Experimental::Internal::RPageSinkDaos::StageClusterImpl ( )
finalprotectedvirtual

Returns the number of bytes written to storage (excluding metadata)

Implements ROOT::Experimental::Internal::RPagePersistentSink.

Definition at line 431 of file RPageStorageDaos.cxx.

◆ WriteNTupleAnchor()

void ROOT::Experimental::Internal::RPageSinkDaos::WriteNTupleAnchor ( )
protected

Definition at line 485 of file RPageStorageDaos.cxx.

◆ WriteNTupleFooter()

void ROOT::Experimental::Internal::RPageSinkDaos::WriteNTupleFooter ( const void *  data,
size_t  nbytes,
size_t  lenFooter 
)
protected

Definition at line 476 of file RPageStorageDaos.cxx.

◆ WriteNTupleHeader()

void ROOT::Experimental::Internal::RPageSinkDaos::WriteNTupleHeader ( const void *  data,
size_t  nbytes,
size_t  lenHeader 
)
protected

Definition at line 467 of file RPageStorageDaos.cxx.

Member Data Documentation

◆ fCageSizeLimit

uint32_t ROOT::Experimental::Internal::RPageSinkDaos::fCageSizeLimit {}
private

Definition at line 119 of file RPageStorageDaos.hxx.

◆ fClusterGroupId

std::atomic<std::uint64_t> ROOT::Experimental::Internal::RPageSinkDaos::fClusterGroupId {0}
private

Cluster group counter for the next committed cluster pagelist; incremented in CommitClusterGroupImpl()

Definition at line 111 of file RPageStorageDaos.hxx.

◆ fDaosContainer

std::unique_ptr<RDaosContainer> ROOT::Experimental::Internal::RPageSinkDaos::fDaosContainer
private

Underlying DAOS container.

An internal std::shared_ptr keep the pool connection alive. ISO C++ ensures the correct destruction order, i.e., ~RDaosContainer is invoked first (which calls daos_cont_close(); the destructor for the std::shared_ptr<RDaosPool> is invoked after (which calls daos_pool_disconect()).

Definition at line 107 of file RPageStorageDaos.hxx.

◆ fNBytesCurrentCluster

std::uint64_t ROOT::Experimental::Internal::RPageSinkDaos::fNBytesCurrentCluster {0}
private

Tracks the number of bytes committed to the current cluster.

Definition at line 115 of file RPageStorageDaos.hxx.

◆ fNTupleAnchor

RDaosNTupleAnchor ROOT::Experimental::Internal::RPageSinkDaos::fNTupleAnchor
private

Definition at line 117 of file RPageStorageDaos.hxx.

◆ fNTupleIndex

ntuple_index_t ROOT::Experimental::Internal::RPageSinkDaos::fNTupleIndex {0}
private

Definition at line 118 of file RPageStorageDaos.hxx.

◆ fPageId

std::atomic<std::uint64_t> ROOT::Experimental::Internal::RPageSinkDaos::fPageId {0}
private

Page identifier for the next committed page; it is automatically incremented in CommitSealedPageImpl()

Definition at line 109 of file RPageStorageDaos.hxx.

◆ fURI

std::string ROOT::Experimental::Internal::RPageSinkDaos::fURI
private

A URI to a DAOS pool of the form 'daos://pool-label/container-label'.

Definition at line 113 of file RPageStorageDaos.hxx.

  • tree/ntuple/v7/inc/ROOT/RPageStorageDaos.hxx
  • tree/ntuple/v7/src/RPageStorageDaos.cxx