org.apache.hadoop.hbase.io.hfile

Interface HFileBlock.BlockIterator

Enclosing class:

HFileBlock

static interface HFileBlock.BlockIterator

Iterator for reading HFileBlocks in load-on-open-section, such as root data index block, meta index block, file info block etc.

Method Summary

Modifier and Type	Method and Description
void	freeBlocks() Now we use the ByteBuffAllocator to manage the nio ByteBuffers for HFileBlocks, so we must deallocate all of the ByteBuffers in the end life.
HFileBlock	nextBlock() Get the next block, or null if there are no more blocks to iterate.
HFileBlock	nextBlockWithBlockType(BlockType blockType) Similar to nextBlock() but checks block type, throws an exception if incorrect, and returns the HFile block

Method Detail

nextBlock

HFileBlock nextBlock()
              throws IOException

Get the next block, or null if there are no more blocks to iterate.

Throws:

IOException

nextBlockWithBlockType

HFileBlock nextBlockWithBlockType(BlockType blockType)
                           throws IOException

Similar to nextBlock() but checks block type, throws an exception if incorrect, and returns the HFile block

Throws:

IOException

freeBlocks

void freeBlocks()

Now we use the ByteBuffAllocator to manage the nio ByteBuffers for HFileBlocks, so we must deallocate all of the ByteBuffers in the end life. the BlockIterator's life cycle is starting from opening an HFileReader and stopped when the HFileReader#close, so we will keep track all the read blocks until we call freeBlocks() when closing the HFileReader. Sum bytes of those blocks in load-on-open section should be quite small, so tracking them should be OK.