@InterfaceAudience.Private public class MemStoreLAB extends Object
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 KeyValues 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
Modifier and Type | Class and Description |
---|---|
static class |
MemStoreLAB.Allocation
The result of a single allocation.
|
Constructor and Description |
---|
MemStoreLAB() |
MemStoreLAB(org.apache.hadoop.conf.Configuration conf,
MemStoreChunkPool pool) |
public MemStoreLAB()
public MemStoreLAB(org.apache.hadoop.conf.Configuration conf, MemStoreChunkPool pool)
public MemStoreLAB.Allocation allocateBytes(int size)
Copyright © 2014 The Apache Software Foundation. All rights reserved.