[
https://issues.apache.org/jira/browse/LUCENE-1483?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12667775#action_12667775
]
Michael McCandless commented on LUCENE-1483:
--------------------------------------------
bq. The only problem with parallelization is that the MultiReaderHitCollector
must be synchronized in some way.
I think we'd have to collect to separate collectors and then merge
(like ParallelMultiSearcher does today)?
I think this (separate thread for the "big" segments, and one thread
for the "long tail") would be a good approach, except I don't like
that the performance would depend so much on the structure of the
index. EG after you've optimized your index you'd suddenly get no
concurrency, and presumably worse performance than when you had a few
big segments.
Could we instead divide the index into chunks and have each thread
skipTo the start of its chunk? EG if the index has N docs, and you
want to use M threads, each thread visits N/M docs. If that can work
it should be less dependent on the index structure.
> Change IndexSearcher multisegment searches to search each individual segment
> using a single HitCollector
> --------------------------------------------------------------------------------------------------------
>
> Key: LUCENE-1483
> URL: https://issues.apache.org/jira/browse/LUCENE-1483
> Project: Lucene - Java
> Issue Type: Improvement
> Affects Versions: 2.9
> Reporter: Mark Miller
> Priority: Minor
> Attachments: LUCENE-1483-backcompat.patch, LUCENE-1483-partial.patch,
> LUCENE-1483.patch, LUCENE-1483.patch, LUCENE-1483.patch, LUCENE-1483.patch,
> LUCENE-1483.patch, LUCENE-1483.patch, LUCENE-1483.patch, LUCENE-1483.patch,
> LUCENE-1483.patch, LUCENE-1483.patch, LUCENE-1483.patch, LUCENE-1483.patch,
> LUCENE-1483.patch, LUCENE-1483.patch, LUCENE-1483.patch, LUCENE-1483.patch,
> LUCENE-1483.patch, LUCENE-1483.patch, LUCENE-1483.patch, LUCENE-1483.patch,
> LUCENE-1483.patch, LUCENE-1483.patch, LUCENE-1483.patch, LUCENE-1483.patch,
> LUCENE-1483.patch, LUCENE-1483.patch, LUCENE-1483.patch, LUCENE-1483.patch,
> LUCENE-1483.patch, LUCENE-1483.patch, LUCENE-1483.patch, LUCENE-1483.patch,
> LUCENE-1483.patch, LUCENE-1483.patch, sortBench.py, sortCollate.py
>
>
> This issue changes how an IndexSearcher searches over multiple segments. The
> current method of searching multiple segments is to use a MultiSegmentReader
> and treat all of the segments as one. This causes filters and FieldCaches to
> be keyed to the MultiReader and makes reopen expensive. If only a few
> segments change, the FieldCache is still loaded for all of them.
> This patch changes things by searching each individual segment one at a time,
> but sharing the HitCollector used across each segment. This allows
> FieldCaches and Filters to be keyed on individual SegmentReaders, making
> reopen much cheaper. FieldCache loading over multiple segments can be much
> faster as well - with the old method, all unique terms for every segment is
> enumerated against each segment - because of the likely logarithmic change in
> terms per segment, this can be very wasteful. Searching individual segments
> avoids this cost. The term/document statistics from the multireader are used
> to score results for each segment.
> When sorting, its more difficult to use a single HitCollector for each sub
> searcher. Ordinals are not comparable across segments. To account for this, a
> new field sort enabled HitCollector is introduced that is able to collect and
> sort across segments (because of its ability to compare ordinals across
> segments). This TopFieldCollector class will collect the values/ordinals for
> a given segment, and upon moving to the next segment, translate any
> ordinals/values so that they can be compared against the values for the new
> segment. This is done lazily.
> All and all, the switch seems to provide numerous performance benefits, in
> both sorted and non sorted search. We were seeing a good loss on indices with
> lots of segments (1000?) and certain queue sizes / queries, but the latest
> results seem to show thats been mostly taken care of (you shouldnt be using
> such a large queue on such a segmented index anyway).
> * Introduces
> ** MultiReaderHitCollector - a HitCollector that can collect across multiple
> IndexReaders. Old HitCollectors are wrapped to support multiple IndexReaders.
> ** TopFieldCollector - a HitCollector that can compare values/ordinals across
> IndexReaders and sort on fields.
> ** FieldValueHitQueue - a Priority queue that is part of the
> TopFieldCollector implementation.
> ** FieldComparator - a new Comparator class that works across IndexReaders.
> Part of the TopFieldCollector implementation.
> ** FieldComparatorSource - new class to allow for custom Comparators.
> * Alters
> ** IndexSearcher uses a single HitCollector to collect hits against each
> individual SegmentReader. All the other changes stem from this ;)
> * Deprecates
> ** TopFieldDocCollector
> ** FieldSortedHitQueue
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
To unsubscribe, e-mail: java-dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: java-dev-h...@lucene.apache.org
