org.apache.cassandra.io.util

Class SegmentedFile

java.lang.Object

org.apache.cassandra.utils.concurrent.SharedCloseableImpl

org.apache.cassandra.io.util.SegmentedFile

All Implemented Interfaces:

java.lang.AutoCloseable, SharedCloseable

Direct Known Subclasses:

BufferedSegmentedFile, CompressedSegmentedFile, MmappedSegmentedFile, PoolingSegmentedFile

public abstract class SegmentedFile
extends SharedCloseableImpl

Abstracts a read-only file that has been split into segments, each of which can be represented by an independent FileDataInput. Allows for iteration over the FileDataInputs, or random access to the FileDataInput for a given position. The JVM can only map up to 2GB at a time, so each segment is at most that size when using mmap i/o. If a segment would need to be longer than 2GB, that segment will not be mmap'd, and a new RandomAccessFile will be created for each access to that segment.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	SegmentedFile.Builder Collects potential segmentation points in an underlying file, and builds a SegmentedFile to represent it.
protected static class	SegmentedFile.Cleanup

Field Summary

Fields

Modifier and Type	Field and Description
ChannelProxy	channel
long	length
long	onDiskLength

Constructor Summary

Constructors

Modifier	Constructor and Description
protected	SegmentedFile(SegmentedFile.Cleanup cleanup, ChannelProxy channel, long length, long onDiskLength)
	SegmentedFile(SegmentedFile copy)

Method Summary

Modifier and Type	Method and Description
long[]	copyReadableBounds() Retrieve the readable bounds if any so they can be cloned into other files such as when downsampling an index summary.
RandomAccessReader	createReader()
RandomAccessReader	createThrottledReader(com.google.common.util.concurrent.RateLimiter limiter)
void	dropPageCache(long before)
static SegmentedFile.Builder	getBuilder(Config.DiskAccessMode mode, boolean compressed)
static SegmentedFile.Builder	getCompressedBuilder(CompressedSequentialWriter writer)
FileDataInput	getSegment(long position)
java.util.Iterator<FileDataInput>	iterator(long position)
java.lang.String	path()
abstract SegmentedFile	sharedCopy()
java.lang.String	toString()

Methods inherited from class org.apache.cassandra.utils.concurrent.SharedCloseableImpl

close, close, isCleanedUp

Methods inherited from class java.lang.Object

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

Field Detail

channel

public final ChannelProxy channel

length

public final long length

onDiskLength

public final long onDiskLength

Constructor Detail

SegmentedFile

protected SegmentedFile(SegmentedFile.Cleanup cleanup,
                        ChannelProxy channel,
                        long length,
                        long onDiskLength)

SegmentedFile

public SegmentedFile(SegmentedFile copy)

Method Detail

path

public java.lang.String path()

sharedCopy

public abstract SegmentedFile sharedCopy()

Returns:

a new instance of the object representing the same state and backed by the same underlying resources. Coordinates with the original (and other instances) when the underlying resource should be closed. Throws an exception if the shared resource has already been closed.

createReader

public RandomAccessReader createReader()

createThrottledReader

public RandomAccessReader createThrottledReader(com.google.common.util.concurrent.RateLimiter limiter)

getSegment

public FileDataInput getSegment(long position)

dropPageCache

public void dropPageCache(long before)

getBuilder

public static SegmentedFile.Builder getBuilder(Config.DiskAccessMode mode,
                                               boolean compressed)

Returns:

A SegmentedFile.Builder.

getCompressedBuilder

public static SegmentedFile.Builder getCompressedBuilder(CompressedSequentialWriter writer)

iterator

public java.util.Iterator<FileDataInput> iterator(long position)

Returns:

An Iterator over segments, beginning with the segment containing the given position: each segment must be closed after use.

copyReadableBounds

public long[] copyReadableBounds()

Retrieve the readable bounds if any so they can be cloned into other files such as when downsampling an index summary. Readable bounds are in between record locations in a file that are good positions for mapping the file such that records don't cross mappings.

toString

public java.lang.String toString()

Overrides:

toString in class java.lang.Object