I didn't follow this thread, but let me point out that certain critical pieces
of code executed during search are synchronized (FileIndexInput or some such).
Thus, with a pile of search threads sharing the same IndexSearcher, this can
cause a bottleneck. In such cases I thin khaving multiple I
I'm sorting the search result using multiple fields. Some of them are
"real" fields, within the index, but one is a "pseudo" field, a custom
sort field, implemented by using a custom SortComparatorSource. I end up
with the sorting conditions in exact this order:
1. Sort by score generate by cu
Hello everyone,
I'm quite new in Lucene stuff and I have a problem. I can't get Lucene Java
working on one of my server. I've setup it on another server for Mediawiki and
it works fine.
It's a GNU/Linux Ubuntu Edgy i686 with kernel 2.6.17-11-server running Apache
2.0 with PHP5 for Mediawiki, s
Hello Erick,
it's an average, definitely.
Best
Hans-Peter
- Original Message -
From: "Erick Erickson" <[EMAIL PROTECTED]>
To:
Sent: Saturday, January 12, 2008 2:20 PM
Subject: Re: Performance problem with IndexWriter
Is this just the first document or is it an average? I can imagi
Is this just the first document or is it an average? I can imagine that
initialization happens at different times under different machines...
Best
Erick
On Jan 12, 2008 6:34 AM, Hans-Peter Stricker <[EMAIL PROTECTED]> wrote:
> Hello,
>
> I have a strange problem: the very same call executes 10 t
What analyzer are you using to index? And is it a different analyzer
when indexing as opposed to searching?
Get a copy of Luke and look in your index to see what is
*actually* indexed. At that point, you can start worrying about
how the query is parsed. Until you know what's in your index,
you're
Hello,
I have a strange problem: the very same call executes 10 times faster under
Windows than under Linux: The line
writer.addDocument(doc)
takes (with the very same documents) < 1ms under Windows, but > 10ms under
Linux. maxBufferedDocs = 1, number of documents to index < 1,
flus
Hi all,
I want to match a string like "ABC#$" with "[EMAIL PROTECTED]@[EMAIL
PROTECTED]@".
The query string generated by QueryParse is as follows:
title:ABC#$*
I
append the "*" so that it becomes a wild card search. It does not
return me any records. I did try escaping '#' and '$'. But th
