[
https://issues.apache.org/jira/browse/JCR-1213?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12544424
]
Ard Schrijvers commented on JCR-1213:
-------------------------------------
"http://jackrabbit.apache.org/doc/arch/operate/index-readers.html "
This is really really nice to read! I just only now understand, the deleted
BitSet idea, and that SharedIndexReader lives during the entire life of a
PersistentIndex. Since in 'older' (existing?) indexes only documents can be
deleted, we keep the same SharedIndexReader even if the index changed, and
keep track of the deleted items in the deleted BitSet. And this should give us
a very good possibility to cache the parent relationships of a node, even
though underlying lucene indexes change because of the deletion of a document
(I am pretty much recapitulating and adding nothing new, and certainly nothing
you don't already know, but see it as thinking out loud :-) )
I'll try to look at the end of the day at your suggestion at the end of your
comment, and also see if i can get a 'hacky' working version with the
readersByBase. A nice solution can be made afterwards if it is a succes. After
reading your documentation I am really confident we can make a very efficient
cache for the docNumbers. I'll investigate as well.... :-)
> UUIDDocId cache does not work properly because of weakReferences in
> combination with new instance for combined indexreader
> ---------------------------------------------------------------------------------------------------------------------------
>
> Key: JCR-1213
> URL: https://issues.apache.org/jira/browse/JCR-1213
> Project: Jackrabbit
> Issue Type: Improvement
> Components: query
> Affects Versions: 1.3.3
> Reporter: Ard Schrijvers
> Fix For: 1.4
>
>
> Queries that use ChildAxisQuery or DescendantSelfAxisQuery make use of
> getParent() functions to know wether the parents are correct and if the
> result is allowed. The getParent() is called recursively for every hit, and
> can become very expensive. Hence, in DocId.UUIDDocId, the parents are cached.
> Currently, docId.UUIDDocId's are cached by having a WeakRefence to the
> CombinedIndexReader, but, this CombinedIndexReader is recreated all the time,
> implying that a gc() is allowed to remove the 'expensive' cache.
> A much better solution is to not have a weakReference to the
> CombinedIndexReader, but to a reference of each indexreader segment. This
> means, that in getParent(int n) in SearchIndex the return
> return id.getDocumentNumber(this) needs to be replaced by return
> id.getDocumentNumber(subReaders[i]); and something similar in
> CachingMultiReader.
> That is all. Obviously, when a node/property is added/removed/changed, some
> parts of the cached DocId.UUIDDocId will be invalid, but mainly small indexes
> are updated frequently, which obviously are less expensive to recompute.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
