[
https://issues.apache.org/jira/browse/LUCENE-2135?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12787624#action_12787624
]
Michael McCandless commented on LUCENE-2135:
--------------------------------------------
bq. all IndexReaders should call purge, mabe put it on toplevel IR.close
default impl?
Hmm... this actually gets tricky to get right, because of the FieldCacheKey.
EG on closing a SegmentReader that's a clone of another, you don't want to
evict it from the FieldCache.
I guess I could fix each of the IndexReader subclasses to evict themselves from
the cache. Let me look into that...
> IndexReader.close should forcefully evict entries from FieldCache
> -----------------------------------------------------------------
>
> Key: LUCENE-2135
> URL: https://issues.apache.org/jira/browse/LUCENE-2135
> Project: Lucene - Java
> Issue Type: Bug
> Components: Index
> Reporter: Michael McCandless
> Assignee: Michael McCandless
> Fix For: 3.1
>
> Attachments: LUCENE-2135.patch
>
>
> Spinoff of java-user thread "heap memory issues when sorting by a string
> field".
> We rely on WeakHashMap to hold our FieldCache, keyed by reader. But this
> lacks immediacy on releasing the reference, after a reader is closed.
> WeakHashMap can't free the key until the reader is no longer referenced by
> the app. And, apparently, WeakHashMap has a further impl detail that requires
> invoking one of its methods for it to notice that a key has just become only
> weakly reachable.
> To fix this, I think on IR.close we should evict entries from the FieldCache,
> as long as the sub-readers are truly closed (refCount dropped to 0).
--
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: [email protected]
For additional commands, e-mail: [email protected]
