lucene-solr-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Otis Gospodnetic (JIRA)" <j...@apache.org>
Subject [jira] Commented: (SOLR-805) DisMax queries are not being cached in QueryResultCache
Date Mon, 27 Oct 2008 17:36:44 GMT

    [ https://issues.apache.org/jira/browse/SOLR-805?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12642968#action_12642968
] 

Otis Gospodnetic commented on SOLR-805:
---------------------------------------

Does this call for 1.3.1 release in the very near future?  Most Solr 1.3 DisMax users will
probably not be able to figure this one out, yet it will hurt them.


> DisMax queries are not being cached in QueryResultCache
> -------------------------------------------------------
>
>                 Key: SOLR-805
>                 URL: https://issues.apache.org/jira/browse/SOLR-805
>             Project: Solr
>          Issue Type: Bug
>          Components: search
>    Affects Versions: 1.3
>         Environment: Using Sun JDK 1.5 and Solr 1.3.0 release on Windows XP
>            Reporter: Todd Feak
>            Priority: Critical
>             Fix For: 1.4
>
>
> I have a DisMax Search Handler set up in my solrconfig.xml to weight results based on
which field a hit was found in. Results seem to be coming back fine, but the exact same query
issued twice will *not* result in a cache hit.
> I have run far enough in the debugger to determine that the hashCode for the BooleanQuery
object is returning a different value each time for the same query. This leads me to believe
there is some random factor involved in it's calculation, such as a default Object hashCode()
implementation somewhere in the chain. Non DisMax queries seem to be caching just fine.
> Where I see this behavior exhibited is on line 47 of the QueryResultKey constructor.
I have not dug in far enough to determine exactly where the hashCode is being incorrectly
calculated. I will try and dig in further tomorrow, but wanted to get some attention on the
bug. 

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message