Package org.apache.hadoop.hbase.regionserver

Class StoreFileReader

java.lang.Object

org.apache.hadoop.hbase.regionserver.StoreFileReader

Direct Known Subclasses:

HalfStoreFileReader

@LimitedPrivate("Phoenix")
@Evolving
public class StoreFileReader
extends Object

Reader for a StoreFile.

Field Summary

Fields

Modifier and Type	Field	Description
private BloomFilterMetrics	bloomFilterMetrics
protected BloomType	bloomFilterType
private boolean	bulkLoadResult
protected org.apache.hadoop.conf.Configuration	conf
private final ReaderContext	context
protected BloomFilter	deleteFamilyBloomFilter
private long	deleteFamilyCnt
protected BloomFilter	generalBloomFilter
private byte[]	lastBloomKey
private KeyValue.KeyOnlyKeyValue	lastBloomKeyOnlyKV
private static final org.slf4j.Logger	LOG
private int	prefixLength
private final HFile.Reader	reader
protected long	sequenceID
private boolean	skipResetSeqId
private final StoreFileInfo	storeFileInfo	All StoreFileReader for the same StoreFile will share the StoreFileInfo.refCount.
protected TimeRange	timeRange

Constructor Summary

Constructors

Modifier	Constructor	Description
(package private)	StoreFileReader()	ONLY USE DEFAULT CONSTRUCTOR FOR UNIT TESTS
private	StoreFileReader(HFile.Reader reader, StoreFileInfo storeFileInfo, ReaderContext context, org.apache.hadoop.conf.Configuration conf)
	StoreFileReader(ReaderContext context, HFileInfo fileInfo, CacheConfig cacheConf, StoreFileInfo storeFileInfo, org.apache.hadoop.conf.Configuration conf)

Method Summary

Modifier and Type	Method	Description
private boolean	checkGeneralBloomFilter(byte[] key, Cell kvKey, BloomFilter bloomFilter)
void	close(boolean evictOnClose)
(package private) void	copyFields(StoreFileReader storeFileReader)
(package private) void	disableBloomFilterForTesting()
BloomType	getBloomFilterType()
CellComparator	getComparator()
long	getDeleteFamilyCnt()
long	getEntries()
long	getFilterEntries()	The number of Bloom filter entries in this store file, or an estimate thereof, if the Bloom filter is not loaded.
Optional<Cell>	getFirstKey()
(package private) BloomFilter	getGeneralBloomFilter()
int	getHFileMinorVersion()
HFile.Reader	getHFileReader()
int	getHFileVersion()
Optional<Cell>	getLastKey()
Optional<byte[]>	getLastRowKey()
long	getMaxTimestamp()
int	getPrefixLength()
ReaderContext	getReaderContext()
(package private) int	getRefCount()	Return the ref count associated with the reader whenever a scanner associated with the reader is opened.
HFileScanner	getScanner(boolean cacheBlocks, boolean pread)	Deprecated. since 2.0.0 and will be removed in 3.0.0.
HFileScanner	getScanner(boolean cacheBlocks, boolean pread, boolean isCompaction)	Deprecated. since 2.0.0 and will be removed in 3.0.0.
long	getSequenceID()
StoreFileScanner	getStoreFileScanner(boolean cacheBlocks, boolean pread, boolean isCompaction, long readPt, long scannerOrder, boolean canOptimizeForNonNullColumn)	Get a scanner to scan over this StoreFile.
long	getTotalBloomSize()
long	getTotalUncompressedBytes()
(package private) long	getUncompressedDataIndexSize()
(package private) void	incrementRefCount()	Indicate that the scanner has started reading with this reader.
long	indexSize()
boolean	isBulkLoaded()
boolean	isPrimaryReplicaReader()
(package private) boolean	isSkipResetSeqId()
long	length()
(package private) void	loadBloomfilter()
void	loadBloomfilter(BlockType blockType, BloomFilterMetrics metrics)
Map<byte[],byte[]>	loadFileInfo()
Optional<Cell>	midKey()
(package private) boolean	passesBloomFilter(Scan scan, SortedSet<byte[]> columns)	Checks whether the given scan passes the Bloom filter (if present).
boolean	passesDeleteFamilyBloomFilter(byte[] row, int rowOffset, int rowLen)
private boolean	passesGeneralRowBloomFilter(byte[] row, int rowOffset, int rowLen)	A method for checking Bloom filters.
boolean	passesGeneralRowColBloomFilter(Cell cell)	A method for checking Bloom filters.
private boolean	passesGeneralRowPrefixBloomFilter(Scan scan)	A method for checking Bloom filters.
boolean	passesKeyRangeFilter(Scan scan)	Checks whether the given scan rowkey range overlaps with the current storefile's
(package private) boolean	passesTimerangeFilter(TimeRange tr, long oldestUnexpiredTS)	Check if this storeFile may contain keys within the TimeRange that have not expired (i.e.
(package private) void	readCompleted()	Indicate that the scanner has finished reading with this reader.
(package private) void	setBloomFilterFaulty(BlockType blockType)
void	setBulkLoaded(boolean bulkLoadResult)
private void	setDeleteFamilyBloomFilterFaulty()
private void	setGeneralBloomFilterFaulty()
void	setSequenceID(long sequenceID)
(package private) void	setSkipResetSeqId(boolean skipResetSeqId)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Details

LOG

private static final org.slf4j.Logger LOG

generalBloomFilter

protected BloomFilter generalBloomFilter

deleteFamilyBloomFilter

protected BloomFilter deleteFamilyBloomFilter

bloomFilterMetrics

private BloomFilterMetrics bloomFilterMetrics

bloomFilterType

protected BloomType bloomFilterType

reader

private final HFile.Reader reader

sequenceID

protected long sequenceID

timeRange

protected TimeRange timeRange

lastBloomKey

private byte[] lastBloomKey

deleteFamilyCnt

private long deleteFamilyCnt

bulkLoadResult

private boolean bulkLoadResult

lastBloomKeyOnlyKV

private KeyValue.KeyOnlyKeyValue lastBloomKeyOnlyKV

skipResetSeqId

private boolean skipResetSeqId

prefixLength

private int prefixLength

conf

protected org.apache.hadoop.conf.Configuration conf

storeFileInfo

private final StoreFileInfo storeFileInfo

All StoreFileReader for the same StoreFile will share the StoreFileInfo.refCount. Counter that is incremented every time a scanner is created on the store file. It is decremented when the scan on the store file is done.

context

private final ReaderContext context

Constructor Details

StoreFileReader

private StoreFileReader(HFile.Reader reader,
 StoreFileInfo storeFileInfo,
 ReaderContext context,
 org.apache.hadoop.conf.Configuration conf)

StoreFileReader

public StoreFileReader(ReaderContext context,
 HFileInfo fileInfo,
 CacheConfig cacheConf,
 StoreFileInfo storeFileInfo,
 org.apache.hadoop.conf.Configuration conf)
                throws IOException

Throws:

IOException

StoreFileReader

@Private
StoreFileReader()

ONLY USE DEFAULT CONSTRUCTOR FOR UNIT TESTS

Method Details

copyFields

void copyFields(StoreFileReader storeFileReader)
         throws IOException

Throws:

IOException

isPrimaryReplicaReader

public boolean isPrimaryReplicaReader()

getComparator

public CellComparator getComparator()

getStoreFileScanner

public StoreFileScanner getStoreFileScanner(boolean cacheBlocks,
 boolean pread,
 boolean isCompaction,
 long readPt,
 long scannerOrder,
 boolean canOptimizeForNonNullColumn)

Get a scanner to scan over this StoreFile.

Parameters:

cacheBlocks - should this scanner cache blocks?

pread - use pread (for highly concurrent small readers)

isCompaction - is scanner being used for compaction?

scannerOrder - Order of this scanner relative to other scanners. See KeyValueScanner.getScannerOrder().

canOptimizeForNonNullColumn - true if we can make sure there is no null column, otherwise false. This is a hint for optimization.

Returns:

a scanner

getRefCount

int getRefCount()

Return the ref count associated with the reader whenever a scanner associated with the reader is opened.

incrementRefCount

void incrementRefCount()

Indicate that the scanner has started reading with this reader. We need to increment the ref count so reader is not close until some object is holding the lock

readCompleted

void readCompleted()

Indicate that the scanner has finished reading with this reader. We need to decrement the ref count, and also, if this is not the common pread reader, we should close it.

getScanner

@Deprecated
public HFileScanner getScanner(boolean cacheBlocks,
 boolean pread)

Deprecated.

since 2.0.0 and will be removed in 3.0.0. Do not write further code which depends on this call. Instead use getStoreFileScanner() which uses the StoreFileScanner class/interface which is the preferred way to scan a store with higher level concepts.

Parameters:

cacheBlocks - should we cache the blocks?

pread - use pread (for concurrent small readers)

Returns:

the underlying HFileScanner

See Also:

• HBASE-15296

getScanner

@Deprecated
public HFileScanner getScanner(boolean cacheBlocks,
 boolean pread,
 boolean isCompaction)

Deprecated.

since 2.0.0 and will be removed in 3.0.0. Do not write further code which depends on this call. Instead use getStoreFileScanner() which uses the StoreFileScanner class/interface which is the preferred way to scan a store with higher level concepts. should we cache the blocks? use pread (for concurrent small readers) is scanner being used for compaction?

Returns:

the underlying HFileScanner

See Also:

• HBASE-15296

close

public void close(boolean evictOnClose)
           throws IOException

Throws:

IOException

passesTimerangeFilter

boolean passesTimerangeFilter(TimeRange tr,
 long oldestUnexpiredTS)

Check if this storeFile may contain keys within the TimeRange that have not expired (i.e. not older than oldestUnexpiredTS).

Parameters:

tr - the timeRange to restrict

oldestUnexpiredTS - the oldest timestamp that is not expired, as determined by the column family's TTL

Returns:

false if queried keys definitely don't exist in this StoreFile

passesBloomFilter

boolean passesBloomFilter(Scan scan,
 SortedSet<byte[]> columns)

Checks whether the given scan passes the Bloom filter (if present). Only checks Bloom filters for single-row or single-row-column scans. Bloom filter checking for multi-gets is implemented as part of the store scanner system (see StoreFileScanner.seek(Cell) and uses the lower-level API passesGeneralRowBloomFilter(byte[], int, int) and passesGeneralRowColBloomFilter(Cell).

Parameters:

scan - the scan specification. Used to determine the row, and to check whether this is a single-row ("get") scan.

columns - the set of columns. Only used for row-column Bloom filters.

Returns:

true if the scan with the given column set passes the Bloom filter, or if the Bloom filter is not applicable for the scan. False if the Bloom filter is applicable and the scan fails it.

passesDeleteFamilyBloomFilter

public boolean passesDeleteFamilyBloomFilter(byte[] row,
 int rowOffset,
 int rowLen)

passesGeneralRowBloomFilter

private boolean passesGeneralRowBloomFilter(byte[] row,
 int rowOffset,
 int rowLen)

A method for checking Bloom filters. Called directly from StoreFileScanner in case of a multi-column query.

Returns:

True if passes

passesGeneralRowColBloomFilter

public boolean passesGeneralRowColBloomFilter(Cell cell)

A method for checking Bloom filters. Called directly from StoreFileScanner in case of a multi-column query. the cell to check if present in BloomFilter

Returns:

True if passes

passesGeneralRowPrefixBloomFilter

private boolean passesGeneralRowPrefixBloomFilter(Scan scan)

A method for checking Bloom filters. Called directly from StoreFileScanner in case of a multi-column query.

Returns:

True if passes

checkGeneralBloomFilter

private boolean checkGeneralBloomFilter(byte[] key,
 Cell kvKey,
 BloomFilter bloomFilter)

passesKeyRangeFilter

public boolean passesKeyRangeFilter(Scan scan)

Checks whether the given scan rowkey range overlaps with the current storefile's

Parameters:

scan - the scan specification. Used to determine the rowkey range.

Returns:

true if there is overlap, false otherwise

loadFileInfo

public Map<byte[],byte[]> loadFileInfo()
                                throws IOException

Throws:

IOException

loadBloomfilter

void loadBloomfilter()

loadBloomfilter

public void loadBloomfilter(BlockType blockType,
 BloomFilterMetrics metrics)

setBloomFilterFaulty

void setBloomFilterFaulty(BlockType blockType)

getFilterEntries

public long getFilterEntries()

The number of Bloom filter entries in this store file, or an estimate thereof, if the Bloom filter is not loaded. This always returns an upper bound of the number of Bloom filter entries.

Returns:

an estimate of the number of Bloom filter entries in this file

setGeneralBloomFilterFaulty

private void setGeneralBloomFilterFaulty()

setDeleteFamilyBloomFilterFaulty

private void setDeleteFamilyBloomFilterFaulty()

getLastKey

public Optional<Cell> getLastKey()

getLastRowKey

public Optional<byte[]> getLastRowKey()

midKey

public Optional<Cell> midKey()
                      throws IOException

Throws:

IOException

length

public long length()

getTotalUncompressedBytes

public long getTotalUncompressedBytes()

getEntries

public long getEntries()

getDeleteFamilyCnt

public long getDeleteFamilyCnt()

getFirstKey

public Optional<Cell> getFirstKey()

indexSize

public long indexSize()

getBloomFilterType

public BloomType getBloomFilterType()

getSequenceID

public long getSequenceID()

setSequenceID

public void setSequenceID(long sequenceID)

setBulkLoaded

public void setBulkLoaded(boolean bulkLoadResult)

isBulkLoaded

public boolean isBulkLoaded()

getGeneralBloomFilter

BloomFilter getGeneralBloomFilter()

getUncompressedDataIndexSize

long getUncompressedDataIndexSize()

getTotalBloomSize

public long getTotalBloomSize()

getHFileVersion

public int getHFileVersion()

getHFileMinorVersion

public int getHFileMinorVersion()

getHFileReader

public HFile.Reader getHFileReader()

disableBloomFilterForTesting

void disableBloomFilterForTesting()

getMaxTimestamp

public long getMaxTimestamp()

isSkipResetSeqId

boolean isSkipResetSeqId()

setSkipResetSeqId

void setSkipResetSeqId(boolean skipResetSeqId)

getPrefixLength

public int getPrefixLength()

getReaderContext

public ReaderContext getReaderContext()