org.apache.lucene.index

Class MergePolicy

java.lang.Object

org.apache.lucene.index.MergePolicy

Direct Known Subclasses:

FilterMergePolicy, LogMergePolicy, NoMergePolicy, TieredMergePolicy

public abstract class MergePolicy
extends Object

Expert: a MergePolicy determines the sequence of primitive merge operations.

Whenever the segments in an index have been altered by IndexWriter, either the addition of a newly flushed segment, addition of many segments from addIndexes* calls, or a previous merge that may now need to cascade, IndexWriter invokes findMerges(org.apache.lucene.index.MergeTrigger, org.apache.lucene.index.SegmentInfos, org.apache.lucene.index.MergePolicy.MergeContext) to give the MergePolicy a chance to pick merges that are now required. This method returns a MergePolicy.MergeSpecification instance describing the set of merges that should be done, or null if no merges are necessary. When IndexWriter.forceMerge is called, it calls findForcedMerges(SegmentInfos, int, Map, MergeContext) and the MergePolicy should then return the necessary merges.

Note that the policy can return more than one merge at a time. In this case, if the writer is using SerialMergeScheduler, the merges will be run sequentially but if it is using ConcurrentMergeScheduler they will be run concurrently.

The default MergePolicy is TieredMergePolicy.

WARNING: This API is experimental and might change in incompatible ways in the next release.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	MergePolicy.MergeAbortedException Thrown when a merge was explicitly aborted because IndexWriter.abortMerges() was called.
static interface	MergePolicy.MergeContext This interface represents the current context of the merge selection process.
static class	MergePolicy.MergeException Exception thrown if there are any problems while executing a merge.
static class	MergePolicy.MergeSpecification A MergeSpecification instance provides the information necessary to perform multiple merges.
static class	MergePolicy.OneMerge OneMerge provides the information necessary to perform an individual primitive merge operation, resulting in a single new segment.
static class	MergePolicy.OneMergeProgress Progress and state for an executing merge.

Field Summary

Fields

Modifier and Type	Field and Description
protected static long	DEFAULT_MAX_CFS_SEGMENT_SIZE Default max segment size in order to use compound file system.
protected static double	DEFAULT_NO_CFS_RATIO Default ratio for compound file system usage.
protected long	maxCFSSegmentSize If the size of the merged segment exceeds this value then it will not use compound file format.
protected double	noCFSRatio If the size of the merge segment exceeds this ratio of the total index size then it will remain in non-compound format

Constructor Summary

Constructors

Modifier	Constructor and Description
	MergePolicy() Creates a new merge policy instance.
protected	MergePolicy(double defaultNoCFSRatio, long defaultMaxCFSSegmentSize) Creates a new merge policy instance with default settings for noCFSRatio and maxCFSSegmentSize.

Method Summary

Modifier and Type	Method and Description
protected boolean	assertDelCount(int delCount, SegmentCommitInfo info) Asserts that the delCount for this SegmentCommitInfo is valid
abstract MergePolicy.MergeSpecification	findForcedDeletesMerges(SegmentInfos segmentInfos, MergePolicy.MergeContext mergeContext) Determine what set of merge operations is necessary in order to expunge all deletes from the index.
abstract MergePolicy.MergeSpecification	findForcedMerges(SegmentInfos segmentInfos, int maxSegmentCount, Map<SegmentCommitInfo,Boolean> segmentsToMerge, MergePolicy.MergeContext mergeContext) Determine what set of merge operations is necessary in order to merge to <= the specified segment count.
MergePolicy.MergeSpecification	findFullFlushMerges(MergeTrigger mergeTrigger, SegmentInfos segmentInfos, MergePolicy.MergeContext mergeContext) Identifies merges that we want to execute (synchronously) on commit.
abstract MergePolicy.MergeSpecification	findMerges(MergeTrigger mergeTrigger, SegmentInfos segmentInfos, MergePolicy.MergeContext mergeContext) Determine what set of merge operations are now necessary on the index.
double	getMaxCFSSegmentSizeMB() Returns the largest size allowed for a compound file segment
double	getNoCFSRatio() Returns current noCFSRatio.
protected boolean	isMerged(SegmentInfos infos, SegmentCommitInfo info, MergePolicy.MergeContext mergeContext) Returns true if this single info is already fully merged (has no pending deletes, is in the same dir as the writer, and matches the current compound file setting
boolean	keepFullyDeletedSegment(IOSupplier<CodecReader> readerIOSupplier) Returns true if the segment represented by the given CodecReader should be keep even if it's fully deleted.
protected void	message(String message, MergePolicy.MergeContext mergeContext) Print a debug message to MergePolicy.MergeContext's infoStream.
int	numDeletesToMerge(SegmentCommitInfo info, int delCount, IOSupplier<CodecReader> readerSupplier) Returns the number of deletes that a merge would claim on the given segment.
protected String	segString(MergePolicy.MergeContext mergeContext, Iterable<SegmentCommitInfo> infos) Builds a String representation of the given SegmentCommitInfo instances
void	setMaxCFSSegmentSizeMB(double v) If a merged segment will be more than this value, leave the segment as non-compound file even if compound file is enabled.
void	setNoCFSRatio(double noCFSRatio) If a merged segment will be more than this percentage of the total size of the index, leave the segment as non-compound file even if compound file is enabled.
protected long	size(SegmentCommitInfo info, MergePolicy.MergeContext mergeContext) Return the byte size of the provided SegmentCommitInfo, pro-rated by percentage of non-deleted documents is set.
boolean	useCompoundFile(SegmentInfos infos, SegmentCommitInfo mergedInfo, MergePolicy.MergeContext mergeContext) Returns true if a new segment (regardless of its origin) should use the compound file format.
protected boolean	verbose(MergePolicy.MergeContext mergeContext) Returns true if the info-stream is in verbose mode

Methods inherited from class java.lang.Object

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

Field Detail

DEFAULT_NO_CFS_RATIO

protected static final double DEFAULT_NO_CFS_RATIO

Default ratio for compound file system usage. Set to 1.0, always use compound file system.

See Also:

Constant Field Values

DEFAULT_MAX_CFS_SEGMENT_SIZE

protected static final long DEFAULT_MAX_CFS_SEGMENT_SIZE

Default max segment size in order to use compound file system. Set to Long.MAX_VALUE.

See Also:

Constant Field Values

noCFSRatio

protected double noCFSRatio

If the size of the merge segment exceeds this ratio of the total index size then it will remain in non-compound format

maxCFSSegmentSize

protected long maxCFSSegmentSize

If the size of the merged segment exceeds this value then it will not use compound file format.

Constructor Detail

MergePolicy

public MergePolicy()

Creates a new merge policy instance.

MergePolicy

protected MergePolicy(double defaultNoCFSRatio,
                      long defaultMaxCFSSegmentSize)

Creates a new merge policy instance with default settings for noCFSRatio and maxCFSSegmentSize. This ctor should be used by subclasses using different defaults than the MergePolicy

Method Detail

findMerges

public abstract MergePolicy.MergeSpecification findMerges(MergeTrigger mergeTrigger,
                                                          SegmentInfos segmentInfos,
                                                          MergePolicy.MergeContext mergeContext)
                                                   throws IOException

Determine what set of merge operations are now necessary on the index. IndexWriter calls this whenever there is a change to the segments. This call is always synchronized on the IndexWriter instance so only one thread at a time will call this method.

Parameters:

mergeTrigger - the event that triggered the merge

segmentInfos - the total set of segments in the index

mergeContext - the IndexWriter to find the merges on

Throws:

IOException

findForcedMerges

public abstract MergePolicy.MergeSpecification findForcedMerges(SegmentInfos segmentInfos,
                                                                int maxSegmentCount,
                                                                Map<SegmentCommitInfo,Boolean> segmentsToMerge,
                                                                MergePolicy.MergeContext mergeContext)
                                                         throws IOException

Determine what set of merge operations is necessary in order to merge to <= the specified segment count. IndexWriter calls this when its IndexWriter.forceMerge(int) method is called. This call is always synchronized on the IndexWriter instance so only one thread at a time will call this method.

Parameters:

segmentInfos - the total set of segments in the index

maxSegmentCount - requested maximum number of segments in the index (currently this is always 1)

segmentsToMerge - contains the specific SegmentInfo instances that must be merged away. This may be a subset of all SegmentInfos. If the value is True for a given SegmentInfo, that means this segment was an original segment present in the to-be-merged index; else, it was a segment produced by a cascaded merge.

mergeContext - the MergeContext to find the merges on

Throws:

IOException

findForcedDeletesMerges

public abstract MergePolicy.MergeSpecification findForcedDeletesMerges(SegmentInfos segmentInfos,
                                                                       MergePolicy.MergeContext mergeContext)
                                                                throws IOException

Determine what set of merge operations is necessary in order to expunge all deletes from the index.

Parameters:

segmentInfos - the total set of segments in the index

mergeContext - the MergeContext to find the merges on

Throws:

IOException

findFullFlushMerges

public MergePolicy.MergeSpecification findFullFlushMerges(MergeTrigger mergeTrigger,
                                                          SegmentInfos segmentInfos,
                                                          MergePolicy.MergeContext mergeContext)
                                                   throws IOException

Identifies merges that we want to execute (synchronously) on commit. By default, this will do no merging on commit. If you implement this method in your MergePolicy you must also set a non-zero timeout using IndexWriterConfig.setMaxFullFlushMergeWaitMillis(long). Any merges returned here will make IndexWriter.commit(), IndexWriter.prepareCommit() or IndexWriter.getReader(boolean, boolean) block until the merges complete or until LiveIndexWriterConfig.getMaxFullFlushMergeWaitMillis() has elapsed. This may be used to merge small segments that have just been flushed, reducing the number of segments in the point in time snapshot. If a merge does not complete in the allotted time, it will continue to execute, and eventually finish and apply to future point in time snapshot, but will not be reflected in the current one. If a MergePolicy.OneMerge in the returned MergePolicy.MergeSpecification includes a segment already included in a registered merge, then IndexWriter.commit() or IndexWriter.prepareCommit() will throw a IllegalStateException. Use MergePolicy.MergeContext.getMergingSegments() to determine which segments are currently registered to merge.

Parameters:

mergeTrigger - the event that triggered the merge (COMMIT or GET_READER).

segmentInfos - the total set of segments in the index (while preparing the commit)

mergeContext - the MergeContext to find the merges on, which should be used to determine which segments are already in a registered merge (see MergePolicy.MergeContext.getMergingSegments()).

Throws:

IOException

useCompoundFile

public boolean useCompoundFile(SegmentInfos infos,
                               SegmentCommitInfo mergedInfo,
                               MergePolicy.MergeContext mergeContext)
                        throws IOException

Returns true if a new segment (regardless of its origin) should use the compound file format. The default implementation returns true iff the size of the given mergedInfo is less or equal to getMaxCFSSegmentSizeMB() and the size is less or equal to the TotalIndexSize * getNoCFSRatio() otherwise false.

Throws:

IOException

size

protected long size(SegmentCommitInfo info,
                    MergePolicy.MergeContext mergeContext)
             throws IOException

Return the byte size of the provided SegmentCommitInfo, pro-rated by percentage of non-deleted documents is set.

Throws:

IOException

assertDelCount

protected final boolean assertDelCount(int delCount,
                                       SegmentCommitInfo info)

Asserts that the delCount for this SegmentCommitInfo is valid

isMerged

protected final boolean isMerged(SegmentInfos infos,
                                 SegmentCommitInfo info,
                                 MergePolicy.MergeContext mergeContext)
                          throws IOException

Returns true if this single info is already fully merged (has no pending deletes, is in the same dir as the writer, and matches the current compound file setting

Throws:

IOException

getNoCFSRatio

public double getNoCFSRatio()

Returns current noCFSRatio.

See Also:

setNoCFSRatio(double)

setNoCFSRatio

public void setNoCFSRatio(double noCFSRatio)

If a merged segment will be more than this percentage of the total size of the index, leave the segment as non-compound file even if compound file is enabled. Set to 1.0 to always use CFS regardless of merge size.

getMaxCFSSegmentSizeMB

public double getMaxCFSSegmentSizeMB()

Returns the largest size allowed for a compound file segment

setMaxCFSSegmentSizeMB

public void setMaxCFSSegmentSizeMB(double v)

If a merged segment will be more than this value, leave the segment as non-compound file even if compound file is enabled. Set this to Double.POSITIVE_INFINITY (default) and noCFSRatio to 1.0 to always use CFS regardless of merge size.

keepFullyDeletedSegment

public boolean keepFullyDeletedSegment(IOSupplier<CodecReader> readerIOSupplier)
                                throws IOException

Returns true if the segment represented by the given CodecReader should be keep even if it's fully deleted. This is useful for testing of for instance if the merge policy implements retention policies for soft deletes.

Throws:

IOException

numDeletesToMerge

public int numDeletesToMerge(SegmentCommitInfo info,
                             int delCount,
                             IOSupplier<CodecReader> readerSupplier)
                      throws IOException

Returns the number of deletes that a merge would claim on the given segment. This method will by default return the sum of the del count on disk and the pending delete count. Yet, subclasses that wrap merge readers might modify this to reflect deletes that are carried over to the target segment in the case of soft deletes. Soft deletes all deletes to survive across merges in order to control when the soft-deleted data is claimed.

Parameters:

info - the segment info that identifies the segment

delCount - the number deleted documents for this segment

readerSupplier - a supplier that allows to obtain a CodecReader for this segment

Throws:

IOException

See Also:

IndexWriter.softUpdateDocument(Term, Iterable, Field...), IndexWriterConfig.setSoftDeletesField(String)

segString

protected final String segString(MergePolicy.MergeContext mergeContext,
                                 Iterable<SegmentCommitInfo> infos)

Builds a String representation of the given SegmentCommitInfo instances

message

protected final void message(String message,
                             MergePolicy.MergeContext mergeContext)

Print a debug message to MergePolicy.MergeContext's infoStream.

verbose

protected final boolean verbose(MergePolicy.MergeContext mergeContext)

Returns true if the info-stream is in verbose mode

See Also:

message(String, MergeContext)