Package org.apache.hadoop.hbase.regionserver

Class MemStoreLABImpl

java.lang.Object

org.apache.hadoop.hbase.regionserver.MemStoreLABImpl

All Implemented Interfaces:

MemStoreLAB

@Private
public class MemStoreLABImpl
extends Object
implements MemStoreLAB

A memstore-local allocation buffer.

The MemStoreLAB is basically a bump-the-pointer allocator that allocates big (2MB) byte[] chunks from and then doles it out to threads that request slices into the array.

The purpose of this class is to combat heap fragmentation in the regionserver. By ensuring that all Cells in a given memstore refer only to large chunks of contiguous memory, we ensure that large blocks get freed up when the memstore is flushed.

Without the MSLAB, the byte array allocated during insertion end up interleaved throughout the heap, and the old generation gets progressively more fragmented until a stop-the-world compacting collection occurs.

TODO: we should probably benchmark whether word-aligning the allocations would provide a performance improvement - probably would speed up the Bytes.toLong/Bytes.toInt calls in KeyValue, but some of those are cached anyway. The chunks created by this MemStoreLAB can get pooled at ChunkCreator. When the Chunk comes from pool, it can be either an on heap or an off heap backed chunk. The chunks, which this MemStoreLAB creates on its own (when no chunk available from pool), those will be always on heap backed.

NOTE:if user requested to work with MSLABs (whether on- or off-heap), in CompactingMemStore ctor, the CompactingMemStore.indexType could only be CompactingMemStore.IndexType.CHUNK_MAP,that is to say the immutable segments using MSLABs are going to use CellChunkMap as their index.

Field Summary

Fields

Modifier and Type	Field	Description
private final ChunkCreator	chunkCreator
(package private) Set<Integer>	chunks
private final AtomicBoolean	closed
private AtomicReference<Chunk>	currChunk
private final int	dataChunkSize
private ReentrantLock	lock
(package private) static final org.slf4j.Logger	LOG
private final int	maxAlloc
private final AtomicBoolean	reclaimed
private final RefCnt	refCnt	Its initial value is 1, so it is one bigger than the current count of open scanners which reading data from this MemStoreLAB.

Fields inherited from interface org.apache.hadoop.hbase.regionserver.MemStoreLAB

CHUNK_POOL_INITIALSIZE_KEY, CHUNK_POOL_MAXSIZE_KEY, CHUNK_SIZE_DEFAULT, CHUNK_SIZE_KEY, INDEX_CHUNK_SIZE_PERCENTAGE_DEFAULT, INDEX_CHUNK_SIZE_PERCENTAGE_KEY, MAX_ALLOC_DEFAULT, MAX_ALLOC_KEY, MSLAB_CLASS_NAME, POOL_INITIAL_SIZE_DEFAULT, POOL_MAX_SIZE_DEFAULT, USEMSLAB_DEFAULT, USEMSLAB_KEY

Constructor Summary

Constructors

Constructor	Description
MemStoreLABImpl()
MemStoreLABImpl(org.apache.hadoop.conf.Configuration conf)

Method Summary

Modifier and Type	Method	Description
void	close()	Close this instance since it won't be used any more, try to put the chunks back to pool
private Cell	copyBBECellInto(ByteBufferExtendedCell cell, int maxAlloc)	Mostly a duplicate of copyCellInto(Cell, int)} done for perf sake.
private static Cell	copyBBECToChunkCell(ByteBufferExtendedCell cell, ByteBuffer buf, int offset, int len)	Clone the passed cell by copying its data into the passed buf and create a cell with a chunkid out of it
Cell	copyCellInto(Cell cell)	Allocates slice in this LAB and copy the passed Cell into this area.
private Cell	copyCellInto(Cell cell, int maxAlloc)
private static Cell	copyToChunkCell(Cell cell, ByteBuffer buf, int offset, int len)	Clone the passed cell by copying its data into the passed buf and create a cell with a chunkid out of it
private static Cell	createChunkCell(ByteBuffer buf, int offset, int len, int tagsLen, long sequenceId)
void	decScannerCount()	Called when closing a scanner on the data of this MemStoreLAB
Cell	forceCopyOfBigCellInto(Cell cell)	When a cell's size is too big (bigger than maxAlloc), copyCellInto does not allocate it on MSLAB.
(package private) Chunk	getCurrentChunk()
Chunk	getNewExternalChunk(int size)
Chunk	getNewExternalChunk(ChunkCreator.ChunkType chunkType)
(package private) Integer	getNumOfChunksReturnedToPool(Set<Integer> chunksId)
private Chunk	getOrMakeChunk()	Get the current chunk, or, if there is no current chunk, allocate a new one from the JVM.
(package private) BlockingQueue<Chunk>	getPooledChunks()
(package private) int	getRefCntValue()
void	incScannerCount()	Called when opening a scanner on the data of this MemStoreLAB
(package private) boolean	isClosed()
boolean	isOffHeap()
boolean	isOnHeap()
(package private) boolean	isReclaimed()
private void	recycleChunks()
private void	tryRetireChunk(Chunk c)	Try to retire the current chunk if it is still c.

Methods inherited from class java.lang.Object

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

Field Details

LOG

static final org.slf4j.Logger LOG

currChunk

private AtomicReference<Chunk> currChunk

lock

private ReentrantLock lock

chunks

Set<Integer> chunks

dataChunkSize

private final int dataChunkSize

maxAlloc

private final int maxAlloc

chunkCreator

private final ChunkCreator chunkCreator

closed

private final AtomicBoolean closed

reclaimed

private final AtomicBoolean reclaimed

refCnt

private final RefCnt refCnt

Its initial value is 1, so it is one bigger than the current count of open scanners which reading data from this MemStoreLAB.

Constructor Details

MemStoreLABImpl

public MemStoreLABImpl()

MemStoreLABImpl

public MemStoreLABImpl(org.apache.hadoop.conf.Configuration conf)

Method Details

copyCellInto

public Cell copyCellInto(Cell cell)

Description copied from interface: MemStoreLAB

Allocates slice in this LAB and copy the passed Cell into this area. Returns new Cell instance over the copied the data. When this MemStoreLAB can not copy this Cell, it returns null.

Specified by:

copyCellInto in interface MemStoreLAB

forceCopyOfBigCellInto

public Cell forceCopyOfBigCellInto(Cell cell)

When a cell's size is too big (bigger than maxAlloc), copyCellInto does not allocate it on MSLAB. Since the process of flattening to CellChunkMap assumes that all cells are allocated on MSLAB, during this process, the big cells are copied into MSLAB using this method.

Specified by:

forceCopyOfBigCellInto in interface MemStoreLAB

copyBBECellInto

private Cell copyBBECellInto(ByteBufferExtendedCell cell,
 int maxAlloc)

Mostly a duplicate of copyCellInto(Cell, int)} done for perf sake. It presumes ByteBufferExtendedCell instead of Cell so we deal with a specific type rather than the super generic Cell. Removes instanceof checks. Shrinkage is enough to make this inline where before it was too big. Uses less CPU. See HBASE-20875 for evidence.

See Also:

• copyCellInto(Cell, int)

copyCellInto

private Cell copyCellInto(Cell cell,
 int maxAlloc)

See Also:

• copyBBECellInto(ByteBufferExtendedCell, int)

copyToChunkCell

private static Cell copyToChunkCell(Cell cell,
 ByteBuffer buf,
 int offset,
 int len)

Clone the passed cell by copying its data into the passed buf and create a cell with a chunkid out of it

See Also:

• copyBBECToChunkCell(ByteBufferExtendedCell, ByteBuffer, int, int)

copyBBECToChunkCell

private static Cell copyBBECToChunkCell(ByteBufferExtendedCell cell,
 ByteBuffer buf,
 int offset,
 int len)

Clone the passed cell by copying its data into the passed buf and create a cell with a chunkid out of it

See Also:

• copyToChunkCell(Cell, ByteBuffer, int, int)

createChunkCell

private static Cell createChunkCell(ByteBuffer buf,
 int offset,
 int len,
 int tagsLen,
 long sequenceId)

close

public void close()

Close this instance since it won't be used any more, try to put the chunks back to pool

Specified by:

close in interface MemStoreLAB

getRefCntValue

int getRefCntValue()

incScannerCount

public void incScannerCount()

Called when opening a scanner on the data of this MemStoreLAB

Specified by:

incScannerCount in interface MemStoreLAB

decScannerCount

public void decScannerCount()

Called when closing a scanner on the data of this MemStoreLAB

Specified by:

decScannerCount in interface MemStoreLAB

recycleChunks

private void recycleChunks()

tryRetireChunk

private void tryRetireChunk(Chunk c)

Try to retire the current chunk if it is still c. Postcondition is that curChunk.get() != c

Parameters:

c - the chunk to retire

getOrMakeChunk

private Chunk getOrMakeChunk()

Get the current chunk, or, if there is no current chunk, allocate a new one from the JVM.

getNewExternalChunk

public Chunk getNewExternalChunk(ChunkCreator.ChunkType chunkType)

Specified by:

getNewExternalChunk in interface MemStoreLAB

getNewExternalChunk

public Chunk getNewExternalChunk(int size)

Specified by:

getNewExternalChunk in interface MemStoreLAB

isOnHeap

public boolean isOnHeap()

Specified by:

isOnHeap in interface MemStoreLAB

isOffHeap

public boolean isOffHeap()

Specified by:

isOffHeap in interface MemStoreLAB

getCurrentChunk

Chunk getCurrentChunk()

getPooledChunks

BlockingQueue<Chunk> getPooledChunks()

getNumOfChunksReturnedToPool

Integer getNumOfChunksReturnedToPool(Set<Integer> chunksId)

isReclaimed

boolean isReclaimed()

isClosed

boolean isClosed()